# Layer2C — AI Infrastructure Decision Intelligence > The CTO Advisor LLC (DBA The Advisor Bench) · thectoadvisor.com · Updated October 8, 2026 Layer2C helps enterprise technology leaders evaluate where AI infrastructure vendors actually operate — across compute, data, orchestration, runtime, reasoning, and applications — and where your organization still owns governance, cost, evidence, and accountability. Each assessment maps vendor capabilities against the 4+1 Layer AI Infrastructure Model and scores every component on portability using the Decision Authority Placement Model (DAPM): Retained, Delegated, or Ceded; every layer also carries a decision-authority reading (Retained, Delegated, Ceded, or Absent): who makes the runtime tradeoff, whether the enterprise can see it, and whether it can override it. Full article: https://thectoadvisor.com/blog/2025/11/05/the-cto-advisor-41-layer-ai-infrastructure-model/ ## Framework **8 Layers:** - Layer 0: Compute & Network Fabric - Layer 1A: Data Storage & Governance - Layer 1B: Context Management & Retrieval - Layer 1C: Data Movement & Pipelines - Layer 2A: Infrastructure Orchestration - Layer 2B: Application Runtime & Execution - Layer 2C: Agentic Infrastructure — The Reasoning Plane - Layer 3 (+1): AI Application Layer **DAPM portability (per component):** - Retained — I possess the capability and can operate it independently of this provider - Delegated — someone else provides the capability, but I can substitute that provider without reconstructing my accumulated opinions - Ceded — changing providers requires reconstructing those opinions **Decision authority (per layer, gaps included):** - Retained — the enterprise, or code it writes or controls, decides - Delegated — the vendor or a model decides; the enterprise can see and override - Ceded — the vendor or a model decides; no override, often invisible - Absent — nothing offered, nothing inherited **Layer Status Indicators:** ● Strength ◑ Delegated/Partial ○ Gap ✕ Absent ◇ Partner Ecosystem ## Vendors Assessed (48 complete) ### AMD Instinct, EPYC & Enterprise AI Suite *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.8 - 4+1 v2: ROCm.ai Tooling GA · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/amd.md **Decision authority (per layer):** Retained 0 · Delegated 4 · Ceded 0 · Absent 4 **Layer Status:** ● Layer 0 · Compute: Compute & Network Fabric — The Second Source — Silicon Across All Three Sub-Layers ○ Layer 1A · Storage: Data Storage & Governance — No Data Foundation — Blueprint, Not Platform ◑ Layer 1B · Retrieval: Context Management & Retrieval — Retrieval Acceleration, Someone Else's Everything Else ○ Layer 1C · Pipelines: Data Movement & Pipelines — Acceleration Primitives Beneath the Pipeline Layer ◑ Layer 2A · Orchestration: Infrastructure Orchestration — MIT-Licensed Orchestration — Governance Without the Licence Line ◑ Layer 2B · Runtime: Application Runtime & Execution — Serving Catalogue on a Standard Endpoint, No Distributed Inference Plane ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Structurally Out of Reach Without a Data Layer ○ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — No Business Applications; Instella Read at 2B **DAPM Portability Profile (components):** Retained 4 · Delegated 8 · Ceded 1 · Absent 0 **Summary:** AMD is a silicon vendor that has climbed into serving without ever building a data layer. Layer 0 is its authority and the only cell where it stands with the frontier: EPYC on the host, Instinct as the accelerator, and Pensando as the fabric, all shipping today. Layers 1B, 2A, 2B, and 3 are real but partial, and all of them rest on the same two artifacts, ROCm and the Enterprise AI Suite. Layers 1A, 1C, and 2C are gaps, the same three NVIDIA and Intel carry, for the same structural reason: an ingredient vendor owns no data foundation, and without one there is nothing for a reasoning plane to reason over. Layer 3 is a gap as well, because AMD sells no business application; its open models are read at 2B. **Validated in Labs:** Lab 017 — Beyond CUDA: the lock was never the silicon (Layer 0, Layer 2B, Layer 2C) — https://labs.layer2c.com/labs/beyond-cuda --- ### Anthropic (Claude Platform + Claude Enterprise + Claude Code) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.5 - 4+1 v2: Stale Follow-Up Cleared · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/anthropic.md **Decision authority (per layer):** Retained 0 · Delegated 1 · Ceded 4 · Absent 3 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Compute Supply Chain, Not Customer Surface ○ Layer 1A · Storage: Data Storage & Governance — Compliance Surface, Not a Governance Catalog ◑ Layer 1B · Retrieval: Context Management & Retrieval — Federated Search, No Retrieval Estate ○ Layer 1C · Pipelines: Data Movement & Pipelines — Feeds and Intake, Not Pipelines ○ Layer 2A · Orchestration: Infrastructure Orchestration — Token Quotas, No Customer Plane ● Layer 2B · Runtime: Application Runtime & Execution — Frontier Serving + Split Agent Runtime ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Runtime Permissions, Not a Reasoning Plane ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — First-Party Value Plane: Work, Code, Cowork **DAPM Portability Profile (components):** Retained 3 · Delegated 4 · Ceded 6 · Absent 0 **Summary:** Anthropic sells intelligence, not infrastructure. The buyer gets frontier Claude models, a serious agent runtime, and three first-party applications: Claude Enterprise for knowledge work, Claude Code for software engineering, and Cowork, which puts the same agentic engine in front of everyone else. Below those surfaces there's nothing to buy. One of the largest compute programs in existence, spanning Amazon's Trainium fleet, up to a million Google TPUs, and 220,000 NVIDIA GPUs rented from SpaceX, sits invisible behind the API. No compute, data platform, pipeline, or orchestration console is for sale. Five of the eight layers stay the enterprise's responsibility, and that's the same shape as OpenAI. The two model vendors are the first rows on this map with identical capability profiles. --- ### Articul8 AI Platform *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.6 - 4+1 v2: Authority Re-Read on Docs · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/articul8.md **Decision authority (per layer):** Retained 0 · Delegated 0 · Ceded 6 · Absent 2 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Enterprise Responsibility ◑ Layer 1A · Storage: Data Storage & Governance — Derived Semantic Layer ● Layer 1B · Retrieval: Context Management & Retrieval — Articul8 Differentiator ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Autonomous Ingestion ○ Layer 2A · Orchestration: Infrastructure Orchestration — Enterprise Responsibility ● Layer 2B · Runtime: Application Runtime & Execution — ModelMesh Runtime ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Governed Mission Dispatch — Probabilistic Core ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Domain Applications **DAPM Portability Profile (components):** Retained 0 · Delegated 0 · Ceded 21 · Absent 0 **Summary:** Articul8 is a domain-specific GenAI platform that enters the buyer conversation at Layers 2B/2C/3 — the execution, reasoning, and application layers — and works downward into the data foundation (L1A/1B/1C). It is the heaviest top-of-stack vendor in the instrument: three strong layers (L1B, L2B, L3), three moderate (L1A, L1C, L2C), and two gaps (L0, L2A). Every layer it provides is Ceded. --- ### AWS AI Infrastructure *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.15 - 4+1 v2: Visibility Settled on Docs · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/aws.md **Decision authority (per layer):** Retained 2 · Delegated 4 · Ceded 2 · Absent 0 **Layer Status:** ● Layer 0 · Compute: Compute & Network Fabric — Custom Silicon Full Stack ● Layer 1A · Storage: Data Storage & Governance — Managed Data Foundation ● Layer 1B · Retrieval: Context Management & Retrieval — Managed Retrieval Portfolio ● Layer 1C · Pipelines: Data Movement & Pipelines — Managed Pipelines (Spark/Airflow Substrates) ● Layer 2A · Orchestration: Infrastructure Orchestration — EKS + Capacity Management ● Layer 2B · Runtime: Application Runtime & Execution — Bedrock + Open Paths ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Intelligence-2C Toolkit; Placement Implicit ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Broadest Ecosystem **DAPM Portability Profile (components):** Retained 7 · Delegated 8 · Ceded 21 · Absent 0 **Summary:** AWS is the first vendor in this assessment series that makes a credible claim across every layer of the 4+1 model — including Layer 2C. The structural difference between AWS and every on-prem vendor (Dell, HPE, VAST) is the direction of authority. On-prem vendors build upward from hardware, attempting to extend authority into orchestration and runtime layers. AWS builds downward from managed services, extending authority into custom silicon (Trainium, Inferentia, Graviton), custom networking (EFA/SRD, Nitro), and now on-prem infrastructure (AWS AI Factories). **Validated in Labs:** Lab 012 — Buy the harness, not the tier (Layer 0) — https://labs.layer2c.com/labs/harness-or-tier; Lab 009 — Recovered capacity is real, and it fails honest (Layer 0) — https://labs.layer2c.com/labs/xeon-rematch; Lab 002 — Own the weights, or the platform owns you (Layer 2B, Layer 2C) — https://labs.layer2c.com/labs/fine-tune-economics; Lab 001 — Borrow the vendor’s plumbing, not its judgment (Layer 1A, Layer 1B) — https://labs.layer2c.com/labs/spark-s3vectors --- ### Microsoft Azure AI Infrastructure *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.15 - 4+1 v2: Layer 2C Re-read on the GA Gate · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/azure.md **Decision authority (per layer):** Retained 1 · Delegated 6 · Ceded 1 · Absent 0 **Layer Status:** ● Layer 0 · Compute: Compute & Network Fabric — Hyperscale Compute Stack ● Layer 1A · Storage: Data Storage & Governance — Unified Data Foundation (Fabric/OneLake) ● Layer 1B · Retrieval: Context Management & Retrieval — Managed Retrieval (AI Search) ● Layer 1C · Pipelines: Data Movement & Pipelines — Managed Pipelines (Fabric/ADF) ● Layer 2A · Orchestration: Infrastructure Orchestration — AKS + Arc Hybrid Orchestration ● Layer 2B · Runtime: Application Runtime & Execution — Foundry + OpenAI Runtime ● Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Complete Intelligence-2C Plane Across Products; Gateway Architect-Wired ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Broadest Enterprise Ecosystem **DAPM Portability Profile (components):** Retained 4 · Delegated 6 · Ceded 16 · Absent 0 **Summary:** Microsoft Azure is the most structurally complex vendor in this assessment series because it operates three distinct authority systems simultaneously: a massive cloud infrastructure platform (Azure), a deeply integrated but newly non-exclusive frontier model partnership (OpenAI), and the largest enterprise software installed base on earth (Microsoft 365, Entra ID, Purview, Fabric). No other assessed vendor straddles all three domains. Google Cloud owns its frontier model outright. AWS partners with model providers at arm's length. Dell, HPE, and VAST operate below the model layer entirely. Microsoft is the only vendor that must coordinate authority across infrastructure, model intelligence, and enterprise application identity — and the April 2026 OpenAI restructuring has made that coordination both more flexible and more visible. --- ### Cerebras Systems (CS-3 and CS-4 Systems + Cerebras Cloud Inference and Training + the Model Zoo) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.4 - 4+1 v2: NetApp Calibration Updated · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/cerebras.md **Decision authority (per layer):** Retained 1 · Delegated 1 · Ceded 2 · Absent 4 **Layer Status:** ● Layer 0 · Compute: Compute & Network Fabric — The Wafer-Scale Engine: CS-3 Systems You Can Own and Cluster, a Cerebras Cloud You Rent by the Hour or the Token, and Cerebras-Operated Data Centers; CS-4 Ships This Quarter; No Networking or Storage of Its Own ○ Layer 1A · Storage: Data Storage & Governance — No Data Foundation: a Files API for Batch Jobs (Private Preview) and the Enterprise's Own S3 Bucket for Uploaded Weights ○ Layer 1B · Retrieval: Context Management & Retrieval — No Retrieval and No Embedding Models: Chat Models Only on the Public Tier; the Documented Vector-Store Integration Is Milvus, Pinecone Appears in an Example ◑ Layer 1C · Pipelines: Data Movement & Pipelines — The Model Zoo's Data-Preparation Pipeline (Apache 2.0: Preprocessing, Four-Stage Deduplication, Read Hooks, Token Generators, Dataloaders) Run by the Enterprise or on the Training Cloud, for Cerebras Training Only; a Batch API in Private Preview; No Movement Product ○ Layer 2A · Orchestration: Infrastructure Orchestration — A Metering Edge and a Provisioned Endpoint: Two-Level Rate Limits, Projects, and Roles; Dedicated Endpoints With No Documented Capacity Controls; Service Tiers in Private Preview; No Customer Plane ● Layer 2B · Runtime: Application Runtime & Execution — The Fastest Production Inference of Open Models Behind an OpenAI-Compatible API, Shared or on Dedicated Wafers, With Tool Use, Structured Outputs, Reasoning, and Prompt Caching; Your Own Fine-Tuned Weights in Private Preview; No Agent Runtime ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Account Controls, Not a Plane: Organizations, Projects, Roles, Rate Limits, Audit Logs; Gateways Are Partners' (Cloudflare AI Gateway, Portkey, Kong, Operant) ○ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — No Application: a Playground, Cookbooks, and Customer Stories; the Value Plane Is the Customer's (Cognition, Lovable, Tavus, GSK) or OpenAI's **DAPM Portability Profile (components):** Retained 2 · Delegated 1 · Ceded 3 · Absent 0 **Summary:** Cerebras is the other AI silicon that reached production, and since May 14, 2026 a public company (Nasdaq: CBRS). The map reads it strong at two layers, moderate at one, and gap at five, the narrowest profile in the batch. Layer 0 is strong on the NVIDIA reading: the Wafer-Scale Engine in CS-3 systems the enterprise can own and cluster, CS-4 shipping this quarter, and Cerebras-operated data centers (165 MW in Mikkeli; AMD Helios racks in the second half of 2026) sold as a Training Cloud by the hour and an Inference Cloud by the token. Layer 2B is strong: the fastest production inference of open-weight models behind an OpenAI-compatible API with tool use, constrained-decoding structured outputs, reasoning controls, and zero-retention prompt caching, on a shared tier of two models or on dedicated wafers with a long catalogue (the enterprise's own fine-tuned weights are a Private Preview); OpenAI's Ultrafast Mode for GPT-5.6 Sol runs here in a limited preview. Layer 1C is moderate on the Model Zoo's open data-preparation pipeline, a fixed-function slice whose only destination is the Cerebras trainer. Everything else is a gap by design: no data foundation, no embeddings or retrieval, no movement product, no customer-facing orchestration plane (limits and budgets on a meter; the scheduling is Cerebras's, so that gap reads Ceded), no governance plane, no application. --- ### Cisco Secure AI Factory with NVIDIA *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.12 - 4+1 v2: Visibility Settled on Docs · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/cisco.md **Decision authority (per layer):** Retained 1 · Delegated 2 · Ceded 5 · Absent 0 **Layer Status:** ● Layer 0 · Compute: Compute & Network Fabric — Networking Strength — Asymmetric ● Layer 1A · Storage: Data Storage & Governance — Partner-Delivered Data Platform ● Layer 1B · Retrieval: Context Management & Retrieval — Partner-Delivered Retrieval ● Layer 1C · Pipelines: Data Movement & Pipelines — Partner-Delivered Pipeline Engines ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Networking-Centric Orchestration ◑ Layer 2B · Runtime: Application Runtime & Execution — Security Layer — Not Runtime ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Security + Identity — Not Yet Governance ◇ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Partner Ecosystem **DAPM Portability Profile (components):** Retained 6 · Delegated 2 · Ceded 24 · Absent 0 **Summary:** Cisco is the only vendor in this assessment series whose AI infrastructure authority is anchored by networking and security rather than compute, storage, or cloud services. Where Dell builds up from servers, HPE from sovereign compute, VAST from storage, and hyperscalers from managed services, Cisco builds outward from the network fabric and security posture that connects everything else. The Secure AI Factory with NVIDIA is a reference architecture, not a vertically integrated platform — and that distinction defines Cisco's entire 4+1 profile. --- ### Cloudera (Anywhere Cloud + Data Services + Cloudera AI) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.4 - 4+1 v2: 1C Reconcile Note Settled · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/cloudera.md **Decision authority (per layer):** Retained 0 · Delegated 3 · Ceded 2 · Absent 3 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Enterprise Responsibility on Every Path; Cloudera Ships Software, Not Substrate ● Layer 1A · Storage: Data Storage & Governance — Open Lakehouse on Iceberg + SDX Governance + Lineage, Everywhere the Data Sits ○ Layer 1B · Retrieval: Context Management & Retrieval — Retrieval Is in Technical Preview: RAG Studio, Semantic Search, and Tool Calling Unscored ● Layer 1C · Pipelines: Data Movement & Pipelines — DataFlow, Streaming, and Data Engineering: Any Source to Any Destination on Every Path ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Platform-Scoped Kubernetes with GPU Quota Pools; Anywhere Cloud Control Plane Launched, Undocumented ● Layer 2B · Runtime: Application Runtime & Execution — AI Inference (NIM, Hugging Face on vLLM, Triton; Any Model; On-Prem and Cloud) + Agent Studio (GA On Premises) ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Data Governance Reaches the Models; Agent Governance Lives Inside Agent Studio ◑ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — AI Visuals and Annotation in Data Visualization; SQL Assistant in Preview **DAPM Portability Profile (components):** Retained 6 · Delegated 6 · Ceded 13 · Absent 0 **Summary:** Cloudera is a data platform that runs wherever the enterprise puts it, and the map reads it strongest at the two layers that were always its own. Layer 1A is strong on an Apache Iceberg lakehouse with the Shared Data Experience (SDX) governing it and Octopai lineage across the estate; Layer 1C is strong on DataFlow (Apache NiFi 2.0 with 450-plus connectors), Streaming (Kafka, Flink), and Data Engineering (Spark, Airflow) on every deployment path; Layer 2B is strong on Cloudera AI Inference serving any model on the customer's own GPUs since February 2026 and an Agent Studio that reached general availability on premises in June 2026. Two layers are moderate (2A on platform-scoped Kubernetes with GPU quota pools; Layer 3 on AI features in Data Visualization), and three are gaps: no substrate (the customer supplies it on every path), no generally available retrieval product (RAG Studio and Semantic Search are Technical Preview), and no reasoning plane beyond Agent Studio's own guardrails and audit. --- ### Cloudflare (Workers AI + AI Gateway + Agents SDK + Developer Platform) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.6 - 4+1 v2: 1C Reconcile Note Settled · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/cloudflare.md **Decision authority (per layer):** Retained 1 · Delegated 4 · Ceded 2 · Absent 1 **Layer Status:** ◑ Layer 0 · Compute: Compute & Network Fabric — A Global Network You Can Buy, CPU Compute You Can Size, GPUs You Can't Touch ◑ Layer 1A · Storage: Data Storage & Governance — S3-Compatible Object Store + Edge Databases + Localization Controls; Catalog in Beta ◑ Layer 1B · Retrieval: Context Management & Retrieval — Vectorize + Open Embedding and Rerank Models; Managed RAG in Beta ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Queues, Workflows, and Object Migration; Streaming Ingest and Iceberg in Beta ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Manual Containers from Durable Objects, Multi-Tenant Dispatch, and Jurisdictions; No Autoscaling, No GPU Plane ● Layer 2B · Runtime: Application Runtime & Execution — Serverless Open-Model Inference + One Gateway to 70+ Models + Durable Agents on Every Edge ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Gateway Over Model Calls + Firewall for LLM Endpoints + Access for MCP Servers; Tool Gateway and Identity in Beta ○ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — A Console for AI Traffic, Not an AI Application; the Agentic-Internet Surfaces Are in Beta **DAPM Portability Profile (components):** Retained 1 · Delegated 6 · Ceded 17 · Absent 0 **Summary:** Cloudflare is the network that became a compute platform and then an agent platform, and the map reads it as strong where the agents run and moderate everywhere they store, move, and get governed. One layer strong (2B: open-weights inference on Cloudflare's own GPUs in 180-plus cities, one gateway to 70-plus models across 12-plus providers, and a generally available durable agent runtime with documented approval gates), six moderate (Layer 0 for a network and CPU compute you can buy with no accelerator you can touch, 1A for an S3-compatible store and edge databases without a catalog, 1B for a vector index and open embedding models with the managed search in beta, 1C for queues and durable workflows without pipelines, 2A for manual containers and multi-tenant dispatch, 2C for a gateway over model calls and a firewall for large language model (LLM) endpoints, with the tool gateway and agent identity in beta), and one gap: no first-party application that runs a business process. --- ### Cohere (Command + Embed + Rerank + North) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.2 - 4+1 v2: 1B Ruling · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/cohere.md **Decision authority (per layer):** Retained 0 · Delegated 0 · Ceded 4 · Absent 4 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Enterprise or Cloud Responsibility; Cohere Runs on Rented Capacity ○ Layer 1A · Storage: Data Storage & Governance — Grounding Connectors, Not a Data Foundation ◑ Layer 1B · Retrieval: Context Management & Retrieval — Frontier Embed and Rerank Models + a Managed Retrieval Bundle, Deployable Anywhere ○ Layer 1C · Pipelines: Data Movement & Pipelines — Parsing and Ingest for Retrieval, Not Pipelines ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Model Vault Capacity Controls; Private Deployments Run on the Enterprise's Orchestrator ● Layer 2B · Runtime: Application Runtime & Execution — Frontier Serving Anywhere + Two Apache 2.0 Exits + North Agents (GA) ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Autonomy Policies, Inherited Access, and an Agent Directory, Not a Governance Plane ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — First-Party Value Plane: North, Private and Sovereign **DAPM Portability Profile (components):** Retained 3 · Delegated 4 · Ceded 11 · Absent 0 **Summary:** Cohere is a model provider that sells its models as software the enterprise runs, and the map shows the consequence. Serving and agents read strong at 2B, the value plane reads strong at Layer 3 on North, retrieval and orchestration read moderate on frontier retrieval models with a managed bundle and on Model Vault's capacity surface, and four layers are gaps everywhere the instrument asks for a substrate or a governance plane the customer administers. The private path (a data center or the customer's own cloud account) is where most of the revenue comes from, so the gaps at Layer 0, 1A, and 1C read Absent rather than Ceded-invisible: the enterprise brought its own GPUs, its own Kubernetes, and its own data governance, and Cohere shipped containers into them. --- ### Confluent, an IBM company (Confluent Cloud + Confluent Platform + Confluent Private Cloud + WarpStream + Confluent Intelligence) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.3 - 4+1 v2: MongoDB 2C Ruling · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/confluent.md **Decision authority (per layer):** Retained 1 · Delegated 2 · Ceded 2 · Absent 3 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Not Confluent's Layer (By Design); Confluent Cloud on the Hyperscalers, WarpStream in Your Account, Platform and Private Cloud on Your Kubernetes ◑ Layer 1A · Storage: Data Storage & Governance — The Event Log as a Governed Foundation: Tiered Topics, Tableflow to Iceberg and Delta in Your Buckets, Schema Registry With Data Contracts, Stream Catalog and Lineage ◑ Layer 1B · Retrieval: Context Management & Retrieval — Federated Vector, Text, and Key Search From Flink Into the Enterprise's Own Stores + Real-Time Context Engine Serving Topic State Over MCP; No Confluent-Owned Vector Index ● Layer 1C · Pipelines: Data Movement & Pipelines — The Heartland: Kafka, 100+ Managed Connectors, Flink SQL and Table API, Kafka Streams, Cluster Linking, and Tableflow on Confluent Cloud; Kafka, Connect, and Flink on Platform, Private Cloud, and WarpStream ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Elastic Clusters and Flink Compute Pools With Autopilot on Confluent Cloud; Confluent for Kubernetes, Confluent Manager for Flink, and the Private Cloud Gateway on Yours; No GPU Plane ◑ Layer 2B · Runtime: Application Runtime & Execution — Streaming Agents on Flink (GA May 2026) Against the Enterprise's Own Model Accounts; ML Functions; No Model Serving in Production (Managed Models Early Access) ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Agent Governance Lives Inside the Streaming Agent Runtime (RBAC, Audit, Replay); Kafka Is the Coordination Bus, Not a Governance Plane; A2A in Open Preview ○ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — No Application; the AI Assistant, Connector Troubleshooting, and Copilot Are Preview and Early Access **DAPM Portability Profile (components):** Retained 7 · Delegated 7 · Ceded 17 · Absent 0 **Summary:** Confluent is the data streaming platform, and since March 17, 2026 it's an IBM company that IBM says will keep operating as a distinct brand. The map reads it as strong at one layer and moderate at four. Layer 1C is the heartland: Apache Kafka, one hundred-plus fully managed connectors, Apache Flink as SQL, Table API, user-defined functions and process table functions, Kafka Streams, Cluster Linking, and Tableflow, on a fully managed cloud, a self-managed platform, a private cloud package, and WarpStream's bring-your-own-cloud diskless Kafka. Layer 1A is moderate: the durable event log with tiered storage, Tableflow materializing topics as Iceberg and Delta Lake tables in the enterprise's own buckets, and Stream Governance (Schema Registry, data contracts, catalog, lineage, a data portal) govern data in motion, not a general estate. Layer 1B is moderate on Flink's federated key, text, and vector search against the enterprise's own stores and on the Real-Time Context Engine serving topic state to agents over the Model Context Protocol. Layer 2A is moderate on elastic clusters, Flink compute pools with Autopilot, and Kubernetes operators for the platform's own components. Layer 2B is moderate: Streaming Agents (generally available May 19, 2026) run event-driven, tool-calling agents on Flink against the enterprise's own model accounts, but Confluent serves no models of its own in production (managed models are Early Access) and the agent surface is four months past general availability with reflection, Agent2Agent, and tool invocation still in preview. Layers 0, 2C, and 3 are gaps: no substrate, no agent-governance plane beyond the runtime's own audit and replay, and no application (the AI Assistant and Copilot are preview and Early Access). --- ### CoreWeave AI Cloud *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.14 - 4+1 v2: Visibility Settled on Docs · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/coreweave.md **Decision authority (per layer):** Retained 0 · Delegated 3 · Ceded 1 · Absent 4 **Layer Status:** ● Layer 0 · Compute: Compute & Network Fabric — NVIDIA-Reference Cloud Compute ◑ Layer 1A · Storage: Data Storage & Governance — Open Default Storage + Dedicated Tier ○ Layer 1B · Retrieval: Context Management & Retrieval — Enterprise-provided ○ Layer 1C · Pipelines: Data Movement & Pipelines — Fast Transport, No Pipeline Layer ● Layer 2A · Orchestration: Infrastructure Orchestration — Open-Substrate GPU Orchestration ● Layer 2B · Runtime: Application Runtime & Execution — Strong — authority splits by what you use ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Gap: Improvement Loop, Not a Reasoning Plane ○ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — No Business Applications; W&B Read at 2B **DAPM Portability Profile (components):** Retained 2 · Delegated 4 · Ceded 8 · Absent 0 **Summary:** CoreWeave is a pure-play NVIDIA GPU cloud whose entire differentiation lives at three layers — the compute fabric (0), infrastructure orchestration (2A), and application runtime (2B) — and which is, unusually for a cloud in this assessment, built deliberately on open substrate. Its orchestration opinions rest on Kubernetes (CKS) and Slurm (SUNK), not a proprietary managed scheduler, and SUNK Anywhere runs those same workflows on non-CoreWeave and on-prem clusters with few configuration changes. That makes CoreWeave's orchestration layer the rare cloud cell that reads as Retained rather than Ceded. --- ### Databricks Data Intelligence Platform *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.12 - 4+1 v2: 2C Strong on a Complete GA Plane · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/databricks.md **Decision authority (per layer):** Retained 0 · Delegated 7 · Ceded 1 · Absent 0 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Not Databricks' Layer (By Design) ● Layer 1A · Storage: Data Storage & Governance — Lakehouse Strength — Open Formats, Captive Governance ● Layer 1B · Retrieval: Context Management & Retrieval — Lakehouse-Native Governed Retrieval ● Layer 1C · Pipelines: Data Movement & Pipelines — Data Engineering Heartland ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Managed Workload Orchestration; Serverless GPU Pre-GA ● Layer 2B · Runtime: Application Runtime & Execution — Mosaic AI Runtime — Serving + Agents (GA) ● Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Complete GA Agent-Governance Plane; Placement Routing Beta ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — First-Party Data Apps + Marketplace Ecosystem **DAPM Portability Profile (components):** Retained 1 · Delegated 6 · Ceded 19 · Absent 0 **Summary:** Databricks is a data + AI lakehouse platform that floats above Layer 0 — it owns no silicon and runs on AWS, Azure, and GCP — and is strong across the entire data and application stack: storage and governance (1A), retrieval (1B), pipelines (1C), runtime (2B), and the value plane (3). It moderates only where it does not own the layer: infrastructure orchestration (2A), where it rents cloud compute and its serverless GPU runtime is still in preview. The reasoning plane (2C) reads strong on a complete GA governance plane, though its placement routing is still Beta. Its authority sits in the opinion layer, never the substrate. --- ### Dataiku (DSS 15 + LLM Mesh + Agent Hub + Govern + Dataiku Cloud) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.0 - 4+1 v2: Authority Split · September 10, 2026 Sources and audit trail: https://layer2c.com/assessment/dataiku.md **Decision authority (per layer):** Retained 0 · Delegated 7 · Ceded 0 · Absent 1 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Not Dataiku's Layer (By Design); Dataiku Cloud on Dataiku's Accounts, Cloud Stacks and Custom Installs on Yours ◑ Layer 1A · Storage: Data Storage & Governance — Catalog, Column-Level Lineage, Quality Rules, and a Feature Store Over Data Stored on the Enterprise's Own Connections ◑ Layer 1B · Retrieval: Context Management & Retrieval — Knowledge Banks Over Ten Vector Stores + Semantic Models + Document-Level Security; No Retrieval Engine of Its Own ● Layer 1C · Pipelines: Data Movement & Pipelines — The Flow: Visual and Code Recipes on Any Engine, Any Source to Any Destination, Scenarios, Lineage; Streaming Experimental ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Elastic AI: Kubernetes Clusters Dataiku Creates and Drives, Execution Configs With GPU Limits, Fleet Manager; No Fair-Share Scheduler ● Layer 2B · Runtime: Application Runtime & Execution — LLM Mesh Gateway + Local Open-Weight Serving on vLLM + Visual, Structured, and Code Agents With Human Approval + API Node Model Serving ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Gateway, Govern Registry With Sign-Off and a Deployment Gate on Bundles, Multi-Agent Orchestration, Observability; No First-Class Agent Identity; Agent Management Due October 2026 ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Agent Hub for Business Users, Three Business Applications, Stories and Dashboards, Cobuild; Forty-Plus Unsupported Solutions Unscored **DAPM Portability Profile (components):** Retained 4 · Delegated 4 · Ceded 31 · Absent 0 **Summary:** Dataiku is a software-only AI platform that sits on top of whatever the enterprise already owns, and the map reads it as strong at the two layers it was built for and at the one it rebuilt itself around: data pipelines (Layer 1C), the runtime that serves models and runs agents (Layer 2B), and the applications business users touch (Layer 3). Governance, retrieval, and orchestration are real but partial: a catalog, lineage, quality rules, and a feature store over data stored elsewhere (1A moderate); Knowledge Banks that orchestrate retrieval across Chroma, Elasticsearch, Pinecone, pgvector, Snowflake Cortex Search, Databricks AI Search, and others without a retrieval engine of Dataiku's own (1B moderate); Kubernetes clusters Dataiku creates and drives on the enterprise's cloud accounts (2A moderate); and an agent-governance plane with four of the five legs the instrument asks for, missing only a first-class agent identity (2C moderate). Layer 0 is a gap by design: Dataiku sells no compute, network, or storage, on any of its three deployment shapes. --- ### Dell AI Factory with NVIDIA *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v2.15 - 4+1 v2: Authority Carrier Named · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/dell.md **Decision authority (per layer):** Retained 4 · Delegated 3 · Ceded 0 · Absent 1 **Layer Status:** ● Layer 0 · Compute: Compute & Network Fabric — Dell Strength ● Layer 1A · Storage: Data Storage & Governance — Dell Strength ◑ Layer 1B · Retrieval: Context Management & Retrieval — Elastic-Powered Retrieval ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Dell + Dataloop ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Rack-Level Management ◑ Layer 2B · Runtime: Application Runtime & Execution — Blueprints + Services ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Enterprise Responsibility ◇ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Partner Ecosystem **DAPM Portability Profile (components):** Retained 2 · Delegated 12 · Ceded 10 · Absent 0 **Summary:** Dell has one of the most credible on-prem AI Factory infrastructure stacks in the market. Its credibility comes from physical infrastructure (Layer 0), storage and data lifecycle integration (Layers 1A/1B/1C, where the Dataloop acquisition gives Dell its first proprietary software asset in the data lifecycle), and ecosystem packaging (Layer 3). From the buyer's seat, the authority profile is just as legible: the enterprise retains commodity compute, cedes its storage and pipeline opinions to Dell, and inherits NVIDIA's judgment for everything GPU-aware above the rack. --- ### Elastic (Elasticsearch + Agent Builder + Security + Observability) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.4 - 4+1 v2: Stale Follow-Up Cleared · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/elastic.md **Decision authority (per layer):** Retained 2 · Delegated 3 · Ceded 2 · Absent 1 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Not Elastic's Layer (By Design) ◑ Layer 1A · Storage: Data Storage & Governance — Tiered Index Store + Access-Control Governance, No Catalog ● Layer 1B · Retrieval: Context Management & Retrieval — Native Hybrid Retrieval: Sparse + Dense + Rerank, Permission-Aware ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Telemetry and Content Ingest Fleet; No CDC, No Lineage ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Platform-Scoped Orchestration (ECE, ECK, Autoscaling); No GPU Plane ◑ Layer 2B · Runtime: Application Runtime & Execution — Agent Builder + Workflows Runtime (GA); Brokered LLMs, No LLM Serving ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Observability + Workflow Orchestration + Registry; No Agent Identity, No Gateway ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — First-Party Security, Observability, and Search Applications; Agentic SOC GA **DAPM Portability Profile (components):** Retained 12 · Delegated 5 · Ceded 22 · Absent 0 **Summary:** Elastic is a retrieval and observability company that shipped an agent runtime on top, and the row reads that way: strong where the buyer's data is searched (Layer 1B) and where the buyer's day is run (Layer 3, Security and Observability), moderate across the data foundation, the pipeline fleet, orchestration, the runtime, and the reasoning plane, and a gap at Layer 0 by design. Authority is two vendors in one row. The free-tier core of Elasticsearch and Kibana, self-run under the GNU Affero General Public License v3 (AGPLv3) option, is the enterprise's to keep: Retained at 1A, 1B, 1C, and Layer 3 on the open-source path, with OpenSearch as the fork. Free-tier Elastic License code the enterprise runs itself (Elastic Agent and Fleet, the EDOT Collector, the content connectors, the ECK operator at Basic) reads Retained on the same possess-and-operate line. Everything AI-shaped on top is paid-tier or proprietary and Ceded: the Elastic Learned Sparse EncodeR (ELSER) and the Jina models, the ranking pipeline, the tiering engine and its snapshot format, Agent Builder, Workflows, the Model Context Protocol (MCP) catalog, Security, and Observability. --- ### Everpure, formerly Pure Storage (Everpure Platform: FlashArray + FlashBlade + Purity + Fusion + Pure1 + Evergreen + Portworx + Data Intelligence + Data Stream) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.3 - 4+1 v2: Overdrive Scored · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/everpure.md **Decision authority (per layer):** Retained 0 · Delegated 1 · Ceded 5 · Absent 2 **Layer Status:** ◑ Layer 0 · Compute: Compute & Network Fabric — Software-Defined All-Flash Storage You Own (FlashArray, FlashBlade//S, //E, and //EXA on DirectFlash) or Subscribe To (Evergreen//One, Cloud Block Store on AWS and Azure); Compute and Fabric Only Through Cisco's FlashStack and NVIDIA's DGX Designs ◑ Layer 1A · Storage: Data Storage & Governance — Purity Data Services Across Block, File, and Object With SafeMode and ActiveCluster; Fusion Fleet Policy; Pure1 Over Sixty Thousand Arrays; Data Intelligence (Formerly 1touch) Announced Available but Undocumented ◑ Layer 1B · Retrieval: Context Management & Retrieval — Data Stream 1.0.0 (August 12, 2026): an On-Appliance Ingest, Embedding, Hybrid Search, and Chat Appliance on FlashBlade With NVIDIA GPUs, Single-Node and 750,000 Pages in Its First Release; Data Intelligence Context Over MCP Announced, Undocumented ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Fleet-Wide Replication and Mobility (ActiveCluster for Block, ActiveDR, CloudSnap, Portworx Disaster Recovery and Migration) and Data Stream's Ingest, Curation, Embedding, and Indexing Pipeline; KV Cache Accelerator, File ActiveCluster, and Workload Rebalance Undocumented or Pending ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Fusion Fleet Automation With Presets, Workloads, and Placement Recommendations; Pure1 Capacity Management; Portworx for Kubernetes and KubeVirt VMs on the Enterprise's Clusters With a Standard CSI Door; No Compute or GPU Scheduling ◑ Layer 2B · Runtime: Application Runtime & Execution — Data Stream's On-Appliance Language Model Behind an OpenAI-Compatible Chat Endpoint, Answering Only Over Its Own Index; the Open-Source Fusion MCP Server for Storage Operations (Pure1 Manage MCP Undocumented); No General Serving, No Agent Runtime ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Attribute-Based Data Policy for Agents Over MCP and a Human-in-the-Loop Gate on Storage Actions; Not a Plane Over Agents ○ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — No Value Plane: Storage Beneath Others' Applications; Copilot Workflow Execution Is a Storage-Operations Assistant With an Unconfirmed Q2 2026 Date; Validated Designs Are Reference Architectures **DAPM Portability Profile (components):** Retained 1 · Delegated 3 · Ceded 14 · Absent 0 **Summary:** Everpure is Pure Storage under a new name (February 23, 2026; ticker P since April 17; S&P 500 on September 21, 2026) with a data-management thesis bolted onto a storage company: the Enterprise Data Cloud, which the company itself calls 'an architectural approach ... not a product'. The map reads it moderate at six layers and gap at two, with no strong cell. Layer 1A is the deepest cell, moderate at the top of the band: Purity's data services across block, file, and object, SafeMode immutability, ActiveCluster and ActiveDR, Fusion's fleet-wide policy enforcement, and Pure1 over sixty thousand arrays; Data Intelligence (the acquired 1touch), announced 'available now' on June 17, 2026 as a catalog across clouds, SaaS, mainframes, and third-party storage, has no documentation on Everpure's support portal and is named, not scored, under rule 3: available now is not GA. Layer 0 is moderate on a software-defined all-flash substrate the enterprise owns or subscribes to; compute and fabric come only through Cisco's FlashStack and NVIDIA's DGX designs on partners' paper. Layer 1B is moderate, on Data Stream 1.0.0 (August 12, 2026): an on-appliance ingestion, embedding, hybrid-search, and chat appliance on FlashBlade with eight NVIDIA GPUs, held to moderate by its first-release limits (one node, 750,000 pages). Layer 1C is moderate on fleet replication and mobility plus Data Stream's preparation pipeline. Layer 2A is moderate on Fusion presets, Pure1 capacity management, and Portworx on the enterprise's Kubernetes, with no compute scheduling. Layer 2B is moderate on Data Stream's OpenAI-compatible chat endpoint over a sealed on-appliance model plus two MCP servers for storage operations; the first draft read it gap on the NetApp precedent. Layer 2C and Layer 3 are gaps: governance of storage actions isn't a plane over agents, and there's no application. --- ### Google Cloud AI Infrastructure *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.19 - 4+1 v2: Layer 1A Re-read on the GA Gate · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/gcp.md **Decision authority (per layer):** Retained 1 · Delegated 3 · Ceded 4 · Absent 0 **Layer Status:** ● Layer 0 · Compute: Compute & Network Fabric — TPU + GPU Full Stack ● Layer 1A · Storage: Data Storage & Governance — Governed Catalog; Gemini Context Graph in Preview ● Layer 1B · Retrieval: Context Management & Retrieval — Model Prep & Managed Retrieval ● Layer 1C · Pipelines: Data Movement & Pipelines — BigQuery-Native Pipelines ● Layer 2A · Orchestration: Infrastructure Orchestration — GKE + Sovereign Extensions ● Layer 2B · Runtime: Application Runtime & Execution — Model-Integrated Stack ● Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Complete Intelligence-2C Plane, No Placement ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Open Model Layer, Captive Platform **DAPM Portability Profile (components):** Retained 4 · Delegated 11 · Ceded 34 · Absent 0 **Summary:** Google Cloud is the only vendor in this assessment series that owns a frontier foundation model — and that single fact restructures the entire 4+1 analysis. Google built Gemini, trains Gemini on its own TPUs, optimizes its silicon for Gemini’s training requirements, and weaves Gemini’s intelligence into every layer of its cloud platform. This creates a model-integrated stack: an architecture where the frontier model is not a component plugged into infrastructure but the intelligence that pervades the infrastructure. **Validated in Labs:** Lab 012 — Buy the harness, not the tier (Layer 0, Layer 2C) — https://labs.layer2c.com/labs/harness-or-tier; Lab 008 — Renting the chip was the easy part (Layer 0) — https://labs.layer2c.com/labs/gemma4-tpu-inference; Lab 007 — The CPU exit is a batch lane, not a serving lane (Layer 0) — https://labs.layer2c.com/labs/gemma4-xeon-inference; Lab 006 — Put the judgment in the constraints, not the weights (Layer 2B) — https://labs.layer2c.com/labs/constraints-not-weights; Lab 004 — You can’t automate a process you haven’t encoded (Layer 1A, Layer 1B, Layer 2B, Layer 2C) — https://labs.layer2c.com/labs/migration-control-plane --- ### Groq (GroqCloud on the LPU + Compound + Server-Side Web Retrieval; Independent Under a Non-Exclusive NVIDIA Technology License) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.1 - 4+1 v2: 1B Authority Reads the Retrieval Call · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/groq.md **Decision authority (per layer):** Retained 0 · Delegated 1 · Ceded 3 · Absent 4 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Not a Customer Surface: Groq's Own LPU Silicon in Thirteen Partner-Operated Data Centers, Sold Only as Tokens and Regional Sovereign Endpoints; a Dedicated Instance Documented Only as a LoRA Deployment Mode; GroqMetal and GroqRack Marketed Without Documentation ○ Layer 1A · Storage: Data Storage & Governance — No Data Foundation: No Retention by Default, Zero Data Retention on Request, Batch Files Kept Thirty Days in US Google Cloud Buckets; Feedback Submissions Kept up to Three Years ◑ Layer 1B · Retrieval: Context Management & Retrieval — Server-Side Web Retrieval in Compound and the Built-In Tools: Web Search on Tavily, Browser Search on Exa, and Visit Website, With Search Results and Citations Returned per Request; No Embedding or Reranking Model, No Index Over the Enterprise's Data ○ Layer 1C · Pipelines: Data Movement & Pipelines — No Pipelines: a Batch API Queues Bulk Requests at Half Price Inside a Processing Window of a Day to a Week ○ Layer 2A · Orchestration: Infrastructure Orchestration — A Metering Edge, Not a Plane: On-Demand, Flex, Performance (Provisioned Throughput With a 99.9 Percent SLA), Auto, and Batch Tiers; Per-Project Limits, Spend Caps, Model Permissions; No Instance You Size ● Layer 2B · Runtime: Application Runtime & Execution — The LPU Inference Cloud: Open Models Behind an OpenAI-Compatible API With Strict Structured Outputs, Reasoning, Whisper Speech to Text, and Prompt Caching, Plus Compound, a Production Agentic System With a Fixed Set of Server-Side Tools; Text to Speech and Vision in Preview; Remote MCP and Connectors in Beta; LoRA Inference on Shared or Dedicated Hardware ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Model Permissions, Project Limits, and Roles Govern Accounts, Not Agents; Compound's Server-Side Tools Take a Per-Request Allow-List and Nothing Else; Remote MCP Approvals and Tool Filters Are Beta ○ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — No Business Application: Groq Chat Is a Demonstration Surface, the Examples Page Holds Groq-Authored Templates and Demos, and the Coding Tools (Factory Droid, OpenCode, Kilo Code, Roo Code, Cline) Are Third-Party and Open-Source With a Groq Key **DAPM Portability Profile (components):** Retained 1 · Delegated 2 · Ceded 2 · Absent 0 **Summary:** Groq is the Language Processing Unit (LPU) inference cloud that licensed its technology to NVIDIA (non-exclusive, December 24, 2025) and kept operating: its founder and president left for NVIDIA, Simon Edwards became chief executive, GroqCloud 'will continue to operate without interruption', and the company has since raised $650 million and a $350 million round at a $3.5 billion valuation, become an NVIDIA Cloud Partner, and announced it will run NVIDIA's Groq 3 LPX with Vera Rubin NVL72 beside its own silicon. The map reads it strong at one layer, moderate at one, and gap at six. Layer 2B is the strength: open models (gpt-oss, Llama, Whisper) behind an OpenAI-compatible API with constrained-decoding structured outputs, reasoning, speech to text, and prompt caching, plus Compound, a production agentic system over a fixed harness of server-side web search, browsing, code execution, and Wolfram Alpha that takes no custom tools, and Low-Rank Adaptation (LoRA) inference of the enterprise's own adapters; text to speech and vision are preview, and the Responses API and every Model Context Protocol (MCP) path are beta. Layer 1B is moderate on the Perplexity reading: Compound's web retrieval on Tavily and Exa returns search results and citations per request, over the web and not the enterprise's data. Layer 0 is a gap by shape: Groq's own silicon in thirteen partner-operated data centers is sold only as tokens, sovereign endpoints, and a dedicated instance for LoRA serving, with GroqMetal marketed and GroqRack undocumented, so the enterprise decides nothing there. Layer 2A is a gap on the OpenAI reading: service tiers, provisioned throughput, per-project limits, spend caps, and model permissions are a metering edge, not a plane. Everything else is a gap: no data foundation, no pipelines, no governance plane, no business application. --- ### HPE Private Cloud AI *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.20 - 4+1 v2: Visibility Settled on Docs · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/hpe.md **Decision authority (per layer):** Retained 1 · Delegated 3 · Ceded 4 · Absent 0 **Layer Status:** ● Layer 0 · Compute: Compute & Network Fabric — HPE Strength ● Layer 1A · Storage: Data Storage & Governance — Owned Storage + Inline Metadata Catalog ◑ Layer 1B · Retrieval: Context Management & Retrieval — Reference Stack + Partner Orchestration ◑ Layer 1C · Pipelines: Data Movement & Pipelines — HPE + Open Source ● Layer 2A · Orchestration: Infrastructure Orchestration — HPE Strength ◑ Layer 2B · Runtime: Application Runtime & Execution — NVIDIA Runtime + Framework Surface ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — IT-Ops Reasoning + Partner AI Orchestration ◇ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Unleash AI Ecosystem **DAPM Portability Profile (components):** Retained 3 · Delegated 7 · Ceded 29 · Absent 0 **Summary:** HPE presents the most structurally interesting comparison to Dell in the 4+1 model because it makes genuine software authority claims that Dell does not. Three capabilities differentiate HPE's architectural position: GreenLake Intelligence (agentic AI mesh using domain-specific LLMs via MCP — HPE-owned Layer 2A/2C for IT operations), the $14B Juniper Networks acquisition (full networking IP stack from silicon to software — owned networking depth that Dell entirely delegates to NVIDIA Spectrum), and the Unleash AI program with Kamiwaza as the chosen Layer 2C orchestration partner (validated in production at Town of Vail). --- ### Hugging Face (Hub + Team and Enterprise Plans + Storage Buckets + Inference Endpoints + Inference Providers + Jobs + Spaces + the Open-Source Libraries) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.1 - 4+1 v2: MongoDB 2C Ruling · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/huggingface.md **Decision authority (per layer):** Retained 1 · Delegated 3 · Ceded 2 · Absent 2 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Not a Customer Surface: GPU, CPU, and TPU Time by the Minute Attached to Jobs, Spaces, ZeroGPU, and Endpoints, on Capacity Hugging Face Rents in the Hyperscalers; No Fabric, No Bare Metal, Nothing You Administer ◑ Layer 1A · Storage: Data Storage & Governance — The Hub as an Artifact Registry: Git Repositories on Xet, S3-Compatible Storage Buckets, Resource Groups, Gating, Audit Logs, SSO, and Storage Regions; No Table, No Policy Below the Repository ◑ Layer 1B · Retrieval: Context Management & Retrieval — Open Embedding and Reranker Models Served Anywhere (Text Embeddings Inference, Sentence Transformers, Inference Endpoints, Inference Providers) and the datasets Library's FAISS and Elasticsearch Indexes; No Managed Retrieval Estate ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Jobs as a Hosted Data-Processing Runner (Schedules, Webhooks, Parallel Runs, Streamed and Mounted Datasets and Buckets) Plus Distilabel and the datasets Library; No Connectors, No CDC, No Lineage ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Inference Endpoints Autoscaling With Replica Bounds and Scale-to-Zero, Instance and Region Pins, Jobs Flavors and Timeouts, Resource-Group Spend Caps; No Scheduler Over a Fleet You Own ◑ Layer 2B · Runtime: Application Runtime & Execution — Serve Any Open Model Anywhere: Inference Endpoints on Three Clouds (vLLM, SGLang, TEI; TGI in Maintenance) and an OpenAI-Compatible Chat Router Across Eighteen Providers; the Agent Libraries Are Experimental or Thin, and There's No Managed Agent Runtime ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Organization Governance for People and Tokens, a Tool Server, and a Place to Upload Agent Traces; Not a Plane Over Agents ○ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — No Business Application: Spaces Host a Million of Other People's Gradio and Docker Apps, the Inference Playground and Data Studio Are Developer Tools, and Gradio Is an Application Framework Rather Than an Application **DAPM Portability Profile (components):** Retained 5 · Delegated 6 · Ceded 10 · Absent 0 **Summary:** Hugging Face is the distribution point for open models and, since September 3, 2026, the subject of an announced NVIDIA acquisition for $12.93 billion that the NVIDIA post says will leave it 'an open platform' with 'NVIDIA compute ... not required'; the agreement hasn't closed and nothing in the documentation has changed, so the row reads Hugging Face as it ships today. The map reads it moderate at four layers and gap at three, with no strong cell: an open-serving platform whose agent, pipeline, and application surfaces are libraries, a runner, and a host rather than products. Layer 2B is the heartland: Inference Endpoints deploy any Hub model on a chosen instance in AWS, Azure, or Google Cloud Platform (GCP) under vLLM, SGLang, or TEI (TGI is in maintenance mode), Inference Providers route chat completions through an OpenAI-compatible API to eighteen partners with policies, failover, and billing at the provider's rates, and the libraries are Apache 2.0 code the enterprise runs; what's missing is a managed agent runtime, and the one full agent harness (smolagents) is experimental by its own documentation, so the cell reads moderate on rule 5 with strong argued. Layer 0 is a gap by shape: GPU, CPU, and TPU time by the minute (Jobs, Spaces, ZeroGPU) is a setting on Hugging Face's products, on capacity it rents in the hyperscalers, with nothing the enterprise administers or decides. Layer 1A is moderate on the Hub as a governed artifact registry (Git repositories on Xet, S3-compatible buckets, resource groups, gating, audit logs, SSO and SCIM, storage regions) that governs repositories, not tables. Layer 1B is moderate on the open embedders, the engine that serves them (TEI, Sentence Transformers), and the datasets library's indexes, with no managed retrieval estate. Layer 1C is moderate on Jobs as a hosted, event-triggered runner for the enterprise's data code plus Distilabel, with no connectors, CDC, or lineage. Layer 2A is moderate on endpoint autoscaling with replica bounds, instance choices, Jobs flavors, and resource-group spend caps. Layer 3 is a gap under the marketplace ruling: Spaces host a million of other people's Gradio and Docker apps and the Playground and Data Studio are developer tools, with no first-party business application. Layer 2C is a gap: a routing and budget gateway over model calls, a trace viewer, and a catalog of agent artifacts are legs, not a plane. --- ### IBM Cloud *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.0 - IBM Cloud as a Cloud, Split from the IBM Row · October 8, 2026 Sources and audit trail: https://layer2c.com/assessment/ibm-cloud.md **Decision authority (per layer):** Retained 1 · Delegated 5 · Ceded 2 · Absent 0 **Layer Status:** ◑ Layer 0 · Compute: Compute & Network Fabric — Three-Vendor GPU Cloud, a Generation Behind the Frontier ● Layer 1A · Storage: Data Storage & Governance — Object Storage + a Catalog Agents Can Query ● Layer 1B · Retrieval: Context Management & Retrieval — Retrieval Kit Assembled from GA Parts: OpenRAG, Milvus, Elasticsearch, Rerank ● Layer 1C · Pipelines: Data Movement & Pipelines — One Integration Plane (Batch, Streaming, Replication) + Event-Driven Code Engine Jobs ● Layer 2A · Orchestration: Infrastructure Orchestration — Managed Kubernetes and OpenShift with Kueue + Serverless GPU Fleets + Satellite ● Layer 2B · Runtime: Application Runtime & Execution — Tool-Calling Serving on watsonx.ai + Open Serving on OpenShift AI; Gateway Not Yet GA ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Orchestrate Control Plane on IBM Cloud; Agent Identity in Preview ◑ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Domain Agents and AI Analytics as IBM Cloud Services; Models and Builders Not Graded **DAPM Portability Profile (components):** Retained 5 · Delegated 11 · Ceded 14 · Absent 0 **Summary:** IBM Cloud is the cloud that sells you the parts. It is strong at the data and runtime layers (storage and catalog, retrieval, pipelines, orchestration, and serving) and moderate at the edges: a GPU fleet a generation behind the frontier at Layer 0, an agent control plane missing first-class agent identity at 2C, and a narrower first-party application estate at Layer 3. Decision authority reads Delegated through most of the stack, Retained in the pipelines where the enterprise's own flows and code decide, and Ceded only at Layer 0 and Layer 3. --- ### IBM (watsonx + IBM Storage + Confluent + Power/Z) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v2.3 - 4+1 v2: NetApp Calibration Updated · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/ibm.md **Decision authority (per layer):** Retained 1 · Delegated 4 · Ceded 3 · Absent 0 **Layer Status:** ◑ Layer 0 · Compute: Compute & Network Fabric — Owned AI Silicon and Systems Beside the Data (z17, Power11, Spyre) + Owned Storage Arrays; No Fabric ● Layer 1A · Storage: Data Storage & Governance — Owned Storage + Catalog with Classification and Lineage ◑ Layer 1B · Retrieval: Context Management & Retrieval — Milvus Inside watsonx.data + Unstructured Data Pipelines for RAG ● Layer 1C · Pipelines: Data Movement & Pipelines — Confluent + DataStage + Lakehouse Federation ● Layer 2A · Orchestration: Infrastructure Orchestration — IBM-Owned GPU Scheduling: Spectrum LSF + Turbonomic ◑ Layer 2B · Runtime: Application Runtime & Execution — watsonx.ai Model Platform + Granite, on the Red Hat Engine ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Agentic Control Plane Generally Available; Agent Identity in Preview ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — IBM Portfolio: watsonx Agents, Maximo, Cognos and Planning Analytics; Red Hat the Infrastructure Focus **DAPM Portability Profile (components):** Retained 4 · Delegated 1 · Ceded 22 · Absent 0 **Summary:** IBM is the enterprise-estate vendor. It owns silicon and systems for running AI next to the data (z17, Power11, Spyre), storage plus a catalog that answers what the data is (watsonx.data intelligence with integrated lineage), streaming and integration (Confluent, DataStage), GPU scheduling (Spectrum LSF, Turbonomic), a model platform (watsonx.ai, Granite), an agent-governance plane complete in every leg but identity (watsonx Orchestrate's control plane), and a first-party application estate (Orchestrate domain agents, Maximo, Cognos, Planning Analytics). It is strong at 1A, 1C, 2A, and Layer 3, and moderate at Layer 0, 1B, 2B, and 2C. Red Hat is its own row: OpenShift is the substrate IBM's on-premises software runs on (on-premises watsonx requires it), so the two rows compose in the common IBM deployment. IBM Cloud is outside this row. --- ### Intel AI Infrastructure Portfolio *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.9 - 4+1 v2: Layer 0 Runtime Rule · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/intel.md **Decision authority (per layer):** Retained 1 · Delegated 1 · Ceded 0 · Absent 6 **Layer Status:** ◑ Layer 0 · Compute: Compute & Network Fabric — CPU Strong, Accelerator Present-but-Thin, No Fabric ○ Layer 1A · Storage: Data Storage & Governance — Substrate Provider, Not a Storage Offering ○ Layer 1B · Retrieval: Context Management & Retrieval — Open Reference Only, Not an Offering ○ Layer 1C · Pipelines: Data Movement & Pipelines — Substrate Provider, Not a Pipeline Offering ○ Layer 2A · Orchestration: Infrastructure Orchestration — Substrate Provider, No Orchestration Offering ◑ Layer 2B · Runtime: Application Runtime & Execution — Open Runtime Tooling, No Managed Service ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — No Control Plane ○ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Silicon Enabler, Not Application Vendor **DAPM Portability Profile (components):** Retained 4 · Delegated 3 · Ceded 0 · Absent 0 **Summary:** Intel is the only ingredient vendor in this instrument. Every other row — Dell, HPE, AWS, Google Cloud, Azure, VAST, VMware, Cisco, IBM, Red Hat, Oracle — is an integrator or operator that assembles Layers 0 through 3 into a system the enterprise buys whole. Intel is the component and open-standards company all of them buy from. Its authority sits entirely at Layer 0, where it is strong on CPU, present but thin on accelerator, and absent on fabric. Above Layer 0 it ships no procurable platform. But it is the open substrate the rest of the stack runs on, and that distinction is the assessment, not a deficiency to apologize for. **Validated in Labs:** Lab 009 — Recovered capacity is real, and it fails honest (Layer 0, Layer 2B) — https://labs.layer2c.com/labs/xeon-rematch --- ### Kamiwaza AI Orchestration Platform *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.10 - 4+1 v2: L3 Visibility on Public Post · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/kamiwaza.md **Decision authority (per layer):** Retained 2 · Delegated 1 · Ceded 2 · Absent 3 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Enterprise Responsibility ◑ Layer 1A · Storage: Data Storage & Governance — Derived Artifact Layer ● Layer 1B · Retrieval: Context Management & Retrieval — Kamiwaza Differentiator ○ Layer 1C · Pipelines: Data Movement & Pipelines — Enterprise Responsibility ○ Layer 2A · Orchestration: Infrastructure Orchestration — Enterprise Responsibility ◑ Layer 2B · Runtime: Application Runtime & Execution — Kamiwaza Runtime ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Deterministic Enforcement, Partial Plane ◑ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Platform + Kaizen **DAPM Portability Profile (components):** Retained 0 · Delegated 3 · Ceded 15 · Absent 0 **Summary:** Kamiwaza is a software-only AI orchestration platform that enters the buyer conversation at Layer 2C and works downward. It is one of only two vendors in the instrument — alongside Articul8 — whose primary product includes the reasoning plane. Where Articul8's Intelligence-2C focuses on mission decomposition and domain-specific agent routing, Kamiwaza's Intelligence-2C focuses on governance: cross-agent authority, relationship-based access control, and policy enforcement at execution time. The enterprise gets production-validated cross-agent governance that no other on-prem vendor provides independently. --- ### Lenovo Hybrid AI Advantage with NVIDIA *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.11 - 4+1 v2: Visibility Settled on Docs · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/lenovo.md **Decision authority (per layer):** Retained 1 · Delegated 3 · Ceded 3 · Absent 1 **Layer Status:** ● Layer 0 · Compute: Compute & Network Fabric — Lenovo Strength ◑ Layer 1A · Storage: Data Storage & Governance — Owned High-End + Partner-Delivered Storage; No Catalog on Lenovo Paper ◑ Layer 1B · Retrieval: Context Management & Retrieval — Validated Retrieval — Partner Stack ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Storage-Native Movement on Lenovo Paper (NetApp ONTAP) ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Fleet Management + Consumption Orchestration ◑ Layer 2B · Runtime: Application Runtime & Execution — Validated Runtimes + Services ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Enterprise Responsibility ◑ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — First-Party Agents + ISV Ecosystem **DAPM Portability Profile (components):** Retained 3 · Delegated 9 · Ceded 12 · Absent 0 **Summary:** Lenovo has one of the most credible on-prem AI infrastructure foundations in the market. Among the on-prem anchors on this instrument, each builds from what it owns: Dell from servers and storage, HPE from sovereign compute and three owned fabrics, Cisco from network silicon and security. Lenovo's anchor is manufacturing scale and thermal engineering — server breadth, Neptune sixth-generation liquid cooling, eight of the top ten public clouds as customers — and it is the only one of the four that owns neither a network fabric nor, until the April 2026 Infinidat close, any opinion-bearing storage software. Above the rack, capability arrives predominantly as other vendors' IP validated and SKU'd on Lenovo paper: ONTAP, Cloudian, WEKA, Red Hat, Nutanix, NVIDIA. --- ### Mistral AI *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.4 - 4+1 v2: Stale Follow-Up Cleared · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/mistral.md **Decision authority (per layer):** Retained 0 · Delegated 1 · Ceded 4 · Absent 3 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Sovereign Fleet, Not Yet a Customer Surface ○ Layer 1A · Storage: Data Storage & Governance — Sovereignty Apparatus, Not a Governance Catalog ◑ Layer 1B · Retrieval: Context Management & Retrieval — Hosted Libraries, Consumed In-Agent ○ Layer 1C · Pipelines: Data Movement & Pipelines — Transform Primitive, No Pipeline ○ Layer 2A · Orchestration: Infrastructure Orchestration — Metering Edge, No Customer Plane (Compute Pending) ● Layer 2B · Runtime: Application Runtime & Execution — Frontier Serving + Owned-Weights Exit, Agent API Pending ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Guardrails and Approvals, Not a Governance Plane ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — First-Party Value Plane: Vibe Work, Vibe Code, Applied AI **DAPM Portability Profile (components):** Retained 4 · Delegated 4 · Ceded 7 · Absent 0 **Summary:** Mistral AI is a frontier lab that sells inference on the industry's standard interface, publishes frontier-class weights under Apache 2.0, and runs a first-party application estate in Vibe with an embedded-engineer applied-AI lane behind it. It is strong at Layer 2B and Layer 3 and a gap everywhere else. It is headquartered in Paris and it sells in the United States. A US inference region since August 2026. Distribution through Microsoft Foundry and Copilot Studio, Bedrock, Vertex, SageMaker and Snowflake. A Palo Alto office, a US general manager who now runs global revenue, and a Seattle-based CMO hired in June. The infrastructure story (Mistral Compute) and the platform story (Studio's Agents, Workflows, Registry and Observability) are both real and both pre-GA. Every cell rests on public documentation. **Validated in Labs:** Lab 013 — The floor is trained, not sized (Layer 2C) — https://labs.layer2c.com/labs/self-host-floor --- ### MongoDB (Atlas + Enterprise Advanced + Community + Voyage AI) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.3 - 4+1 v2: MongoDB 2C Ruling · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/mongodb.md **Decision authority (per layer):** Retained 1 · Delegated 3 · Ceded 2 · Absent 2 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Not MongoDB's Layer (By Design); Atlas Rides the Hyperscalers, Enterprise Advanced Rides Yours ◑ Layer 1A · Storage: Data Storage & Governance — Operational System of Record + Access Governance; No Catalog, No Lineage ● Layer 1B · Retrieval: Context Management & Retrieval — Native Hybrid Retrieval + Voyage Models Inside the Database, Managed or Self-Run ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Streams and CDC Into and Out of MongoDB; No Lineage, Fixed Destinations ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Cluster Autoscaling, Operators, and Ops Manager; No GPU Plane ◑ Layer 2B · Runtime: Application Runtime & Execution — Embedding Inference + MCP Tool Surface + Triggers; No LLM Serving, No Agent Runtime ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Governed Agent Access to MongoDB (App Connections, MCP Configurations), Not a Reasoning Plane ◑ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Analytics Over MongoDB Data; No Business-Process Applications **DAPM Portability Profile (components):** Retained 6 · Delegated 2 · Ceded 16 · Absent 0 **Summary:** MongoDB is the operational database with the retrieval layer built in, and the map reads it as a data platform whose AI story stops at the store. One layer strong (1B: native hybrid retrieval plus the Voyage models, managed or self-run), five moderate (1A as a system of record without a catalog, 1C for streams and change data capture (CDC) into and out of MongoDB, 2A for cluster autoscaling and operators, 2B for embedding inference and a Model Context Protocol (MCP) tool surface with no model serving or agent runtime, Layer 3 for Charts analytics over MongoDB data), and two gaps where the substrate is a hyperscaler's or the enterprise's own and the reasoning plane is whoever's agent platform calls MongoDB. --- ### Nebius (Nebius AI Cloud: GPU Compute, Managed Kubernetes and Soperator, Storage, Virtual Networks + Token Factory Inference and Fine-Tuning + the Applications Marketplace) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.2 - 4+1 v2: Layer 0 Runtime Rule · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/nebius.md **Decision authority (per layer):** Retained 0 · Delegated 4 · Ceded 2 · Absent 2 **Layer Status:** ● Layer 0 · Compute: Compute & Network Fabric — An NVIDIA GPU Cloud You Rent by the Second: H100 to B300 Virtual Machines and InfiniBand GPU Clusters With Partition-Key Isolation and Topology Labels in Six Public Regions Across Finland, France, Israel, the UK, and the US; Virtual Networks and Security Groups; No Silicon of Its Own ◑ Layer 1A · Storage: Data Storage & Governance — S3-Compatible Object Storage With Bucket Policies by Key and Prefix, Shared Filesystems Over virtiofs, and a Tenant-Project Identity Model With SAML Single Sign-On, OIDC Workload Credentials, KMS-Backed Encryption, a HIPAA Guideline, and Audit Logs in Preview; No Catalog, No Classification, No Lineage ◑ Layer 1B · Retrieval: Context Management & Retrieval — Embedding and Rerank Endpoints on Open Models in Token Factory, Qdrant as a Managed Application, and Managed PostgreSQL With pgvector; No Retrieval Engine of Nebius's Own ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Object Storage Data Transfers: Continuous or One-Shot Movement Between Buckets, Third-Party S3 Stores, and Azure Blob With Overwrite and Delete Semantics; Data Lab Prepares Fine-Tuning Datasets by SQL Over Inference Logs and S3 Sources; No ETL, Lineage, or Cache Tiering ● Layer 2A · Orchestration: Infrastructure Orchestration — Managed Kubernetes With GPU Node Groups, Operators, a Cluster Autoscaler, and Topology-Aware Scheduling; Managed Soperator (Open-Source Slurm on Kubernetes) With Ephemeral Nodes That Move Capacity Between Training and Inference; Serverless AI Jobs and Endpoints; Quotas, Preemptible Capacity, and Commitments ● Layer 2B · Runtime: Application Runtime & Execution — Token Factory: Open Models Behind OpenAI-Compatible Chat, Responses, Embeddings, Vision, and Rerank Endpoints, Serverless per Token or on Dedicated Endpoints With Replica Bounds, Plus Full or LoRA Fine-Tuning and Custom Speculator Training With Data Lab; Custom Weights and Sandboxes in Beta; Agents Through Managed Open-Source Applications ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Tenant, Project, and Service-Account Identity for People and Automation, Audit Logs in Preview; Nothing Over Agents: Nebius Echo Is a Console Assistant, Sandboxes Are Beta, the Agents Blueprint Is a Reference Architecture ○ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — No Business Application: an Applications Marketplace of Others' Open-Source Apps Deployed as Managed Services, Onto the Enterprise's Kubernetes, or Onto a Root-Access Virtual Machine, One NVIDIA Blueprint Under a Bring-Your-Own License, and Nebius Echo, a Console Assistant **DAPM Portability Profile (components):** Retained 2 · Delegated 9 · Ceded 9 · Absent 0 **Summary:** Nebius is the European NVIDIA cloud with a model service on top. The map reads it strong at three layers, moderate at three, and gap at two. Layer 0 is strong on GPU virtual machines and InfiniBand GPU clusters from H100 to B300 in six public regions across Finland, France, Israel, the UK, and the US (three private regions serve existing customers), with partition-key isolation and topology labels, virtual networks, and security groups; the capacity is Nebius's own and, by contract, partners' too. Layer 2A is strong on Managed Kubernetes with NVIDIA's operators and topology-aware scheduling, Managed Soperator (its open-source Slurm-on-Kubernetes operator) with ephemeral nodes that move capacity between training and inference, Serverless AI, and quotas, the CoreWeave scheduler pair. Layer 2B is strong on Token Factory: open models behind OpenAI-compatible chat, responses, vision, embeddings, and rerank endpoints, serverless per token or on dedicated endpoints with replica bounds, full or Low-Rank Adaptation (LoRA) fine-tuning and custom speculator training with Data Lab and export to Hugging Face format, and open-source agent tooling as managed applications; custom weights and Sandboxes are beta. Layer 1A is moderate on S3-compatible object storage with bucket policies by key and prefix, virtiofs filesystems, and a tenant-project identity model with Security Assertion Markup Language (SAML) single sign-on and OpenID Connect (OIDC) workload credentials, with audit logs in preview and no catalog. Layer 1B is moderate on embedding and rerank endpoints, Qdrant as a managed app, and Managed PostgreSQL with pgvector. Layer 1C is moderate on Object Storage data transfers (a managed sync between buckets, third-party S3 stores, and Azure Blob) and Data Lab's captive dataset workbench. Layer 3 is a gap under the marketplace ruling: a marketplace of other people's applications, one NVIDIA blueprint under its own license, and Nebius Echo, a console assistant with approval-gated actions, none of them a business application. Layer 2C is a gap: identity for people and workloads, nothing generally available over agents. --- ### NetApp Intelligent Data Infrastructure *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.13 - 4+1 v2: 1B Authority Reads Code · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/netapp.md **Decision authority (per layer):** Retained 2 · Delegated 0 · Ceded 4 · Absent 2 **Layer Status:** ◑ Layer 0 · Compute: Compute & Network Fabric — Software-Defined Storage Fabric ◑ Layer 1A · Storage: Data Storage & Governance — Data Foundation; AI-Metadata Catalog in Private Preview ◑ Layer 1B · Retrieval: Context Management & Retrieval — Validated RAG Design (AIPod Mini); Native Retrieval on the Roadmap ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Movement and Tiering ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Data-Infrastructure Control Plane; No GPU Scheduling ○ Layer 2B · Runtime: Application Runtime & Execution — Runtime is NVIDIA's (Meet-in-the-Middle) ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — No Reasoning Plane (Data Guardrails ≠ Agent Governance) ○ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — No Value Plane — Data Foundation Beneath Others' Apps **DAPM Portability Profile (components):** Retained 1 · Delegated 3 · Ceded 7 · Absent 0 **Summary:** NetApp is a storage and data-management vendor whose authority concentrates in one of the most mature governed data foundations in the market (Layer 1A) and the storage-adjacent infrastructure around it — a software-defined storage fabric (Layer 0), best-in-class data movement (Layer 1C), and a hybrid-multicloud data control plane (Layer 2A). For the AI execution, reasoning, and application layers, NetApp is the data half of a meet-in-the-middle AI factory (AIPod): the compute, model-serving runtime, retrieval, and applications are NVIDIA's, sold alongside through the channel — not NetApp's. --- ### Nutanix Cloud Platform with Enterprise AI *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.9 - 4+1 v2: vGPU Reason Corrected · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/nutanix.md **Decision authority (per layer):** Retained 0 · Delegated 6 · Ceded 1 · Absent 1 **Layer Status:** ◑ Layer 0 · Compute: Compute & Network Fabric — Hardware-Agnostic Abstraction ◑ Layer 1A · Storage: Data Storage & Governance — Platform Storage + Compliance Governance, Not AI-Native ◑ Layer 1B · Retrieval: Context Management & Retrieval — Foundational RAG via Managed Postgres ○ Layer 1C · Pipelines: Data Movement & Pipelines — No Pipeline Layer ● Layer 2A · Orchestration: Infrastructure Orchestration — Nutanix Heritage Strength ◑ Layer 2B · Runtime: Application Runtime & Execution — Platform-Native Serving, Agents Pre-GA ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Gateway Governance, No Placement ◇ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — ISV Ecosystem on a Nutanix Platform **DAPM Portability Profile (components):** Retained 2 · Delegated 9 · Ceded 9 · Absent 0 **Summary:** Nutanix occupies the same structural position as VMware — an HCI / private-cloud platform, an abstraction layer running on commodity x86 it does not manufacture — and against the 4+1 model it produces a strikingly similar profile: strongest at infrastructure orchestration (Layer 2A), competent across the data and runtime layers, and gapped at data pipelines (Layer 1C) and the reasoning plane (Layer 2C). Its deepest IP and authority sit in the orchestration control plane: Prism, AOS, and AHV. --- ### NVIDIA AI Platform *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.16 - 4+1 v2: L0 Authority on the Fabric · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/nvidia.md **Decision authority (per layer):** Retained 0 · Delegated 2 · Ceded 2 · Absent 4 **Layer Status:** ● Layer 0 · Compute: Compute & Network Fabric — NVIDIA Strength — Silicon Authority ○ Layer 1A · Storage: Data Storage & Governance — Accelerator Only ◑ Layer 1B · Retrieval: Context Management & Retrieval — Acceleration + Model Enablement ○ Layer 1C · Pipelines: Data Movement & Pipelines — Gap — single pre-GA capability ● Layer 2A · Orchestration: Infrastructure Orchestration — NVIDIA Authority via Run:ai ● Layer 2B · Runtime: Application Runtime & Execution — NVIDIA Authority — Inference + Agent Runtime ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Runtime Governance Only — Not a Reasoning Plane ○ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — No Business Applications; Nemotron Read at 2B **DAPM Portability Profile (components):** Retained 4 · Delegated 3 · Ceded 9 · Absent 0 **Summary:** NVIDIA is the only vendor in this assessment series that appears inside every other vendor's assessment. Dell's Layer 2A GPU orchestration is NVIDIA Run:ai. HPE's Layer 2B runtime is NVIDIA AI Enterprise. VMware's GPU integration is NVIDIA vGPU Manager. AWS, Google Cloud, and Azure all run NVIDIA GPUs alongside their own silicon. VAST embeds NVIDIA GPUs, NICs, DPUs, and switches into its data platform. Mapping NVIDIA as a standalone vendor inverts the perspective: instead of asking 'where does Dell cede authority to NVIDIA,' the question becomes 'where does NVIDIA claim authority, and from whom?' **Validated in Labs:** Lab 017 — Beyond CUDA: the lock was never the silicon (Layer 0, Layer 2B) — https://labs.layer2c.com/labs/beyond-cuda; Lab 013 — The floor is trained, not sized (Layer 0) — https://labs.layer2c.com/labs/self-host-floor; Lab 012 — Buy the harness, not the tier (Layer 0, Layer 2C) — https://labs.layer2c.com/labs/harness-or-tier; Lab 011 — Buy the labor, not the judgment (Layer 0, Layer 2C) — https://labs.layer2c.com/labs/labor-not-judgment; Lab 009 — Recovered capacity is real, and it fails honest (Layer 0, Layer 2B, Layer 2C) — https://labs.layer2c.com/labs/xeon-rematch; Lab 006 — Put the judgment in the constraints, not the weights (Layer 0, Layer 2B) — https://labs.layer2c.com/labs/constraints-not-weights; Lab 004 — You can’t automate a process you haven’t encoded (Layer 0, Layer 2C) — https://labs.layer2c.com/labs/migration-control-plane; Lab 003 — The validator determines done, not the loop (Layer 0, Layer 2C) — https://labs.layer2c.com/labs/loop-control; Lab 002 — Own the weights, or the platform owns you (Layer 0, Layer 2B, Layer 2C) — https://labs.layer2c.com/labs/fine-tune-economics; Lab 001 — Borrow the vendor’s plumbing, not its judgment (Layer 0, Layer 2B, Layer 2C) — https://labs.layer2c.com/labs/spark-s3vectors --- ### Oracle Cloud Infrastructure (OCI) AI Infrastructure *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.16 - 4+1 v2: Visibility Settled on Docs · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/oci.md **Decision authority (per layer):** Retained 1 · Delegated 4 · Ceded 3 · Absent 0 **Layer Status:** ● Layer 0 · Compute: Compute & Network Fabric — Full-Stack Cloud Compute ● Layer 1A · Storage: Data Storage & Governance — Database-Anchored Data Foundation ● Layer 1B · Retrieval: Context Management & Retrieval — Database-Native Retrieval ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Database-Centric Pipelines ● Layer 2A · Orchestration: Infrastructure Orchestration — OKE + Karpenter + Cluster Networks ● Layer 2B · Runtime: Application Runtime & Execution — OCI Enterprise AI Platform ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Intelligence 2C: Emerging | Infra 2C: Implicit ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Strongest Application-Layer Authority **DAPM Portability Profile (components):** Retained 3 · Delegated 4 · Ceded 23 · Absent 0 **Summary:** OCI occupies a structurally unique position in this assessment series: it is the only hyperscaler whose AI infrastructure strategy is anchored by a database franchise. AWS builds down from managed services. Google builds out from a frontier model. OCI builds up from the enterprise data layer — Oracle AI Database 26ai, Autonomous AI Database, and the Fusion Applications estate that runs 97% of the Fortune 100. Every other hyperscaler treats the database as one service among many. Oracle treats the database as the gravitational center around which AI infrastructure orbits. --- ### OpenAI (API Platform + ChatGPT Enterprise + Codex) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.5 - 4+1 v2: Authority Split · September 2, 2026 Sources and audit trail: https://layer2c.com/assessment/openai.md **Decision authority (per layer):** Retained 0 · Delegated 1 · Ceded 4 · Absent 3 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Not a Customer Surface (By Design) ○ Layer 1A · Storage: Data Storage & Governance — Trust Apparatus, Not a Data Foundation ◑ Layer 1B · Retrieval: Context Management & Retrieval — Hosted RAG Primitive ○ Layer 1C · Pipelines: Data Movement & Pipelines — Intake, Not Movement ○ Layer 2A · Orchestration: Infrastructure Orchestration — Metered Consumption, No Customer Plane ● Layer 2B · Runtime: Application Runtime & Execution — Frontier Serving + Agent Runtime ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Enterprise Responsibility (Frontier Pre-GA) ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — First-Party Value Plane, Assistant Altitude **DAPM Portability Profile (components):** Retained 4 · Delegated 2 · Ceded 9 · Absent 0 **Summary:** OpenAI is the inversion of the traditional infrastructure vendor. The company conducting the largest AI infrastructure buildout in history sells none of it: Stargate, the serving fleet spread across OpenAI-operated, Azure, AWS, and Oracle capacity, and the Jalapeño inference chip are OpenAI's supply chain, not the customer's menu. What a customer can actually deploy is the top of the stack: frontier model serving, an agent runtime, and the most widely used AI application on earth. Everything below that, compute, data foundation, pipelines, orchestration, stays the enterprise's responsibility, by design. --- ### Palantir AIP + Foundry + Apollo *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v3.5 - 4+1 v2: 1C Authority Reads Code · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/palantir.md **Decision authority (per layer):** Retained 2 · Delegated 1 · Ceded 5 · Absent 0 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Not Palantir's Layer (By Design) ● Layer 1A · Storage: Data Storage & Governance — Palantir Strength — Governance Authority, Open Storage ● Layer 1B · Retrieval: Context Management & Retrieval — Palantir Strength — Governed Retrieval ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Foundry Pipelines + Open Compute (MMDP) ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Platform-Scoped Orchestration ● Layer 2B · Runtime: Application Runtime & Execution — Palantir Strength — Governed Agent Runtime ● Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Closest Productized Mechanism ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Palantir Strength — Where It Starts, Not Where It Ends **DAPM Portability Profile (components):** Retained 0 · Delegated 3 · Ceded 26 · Absent 0 **Summary:** Palantir is the first vendor in this series that is not an infrastructure vendor at all, and reading its own documentation makes the inversion precise. Dell, HPE, VAST, Cisco, and the clouds build upward from silicon, storage, or a data center; Palantir builds downward from the decision. Its Architecture Center is explicit that the Ontology is designed to represent 'the complex, interconnected decisions of an enterprise, not simply the data.' Where Dell is strongest at Layer 0 and absent at Layer 2C, Palantir is structurally absent at Layer 0 and strongest at Layers 1A (governance), 2B (governed agent runtime), 2C (reasoning/orchestration), and 3 (value). It is the mirror image of every infrastructure vendor assessed so far. --- ### Perplexity (API Platform + Enterprise + Comet + Computer) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.6 - 4+1 v2: Authority Carrier Named · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/perplexity.md **Decision authority (per layer):** Retained 0 · Delegated 0 · Ceded 5 · Absent 3 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Serving Fleet, Not a Customer Surface ○ Layer 1A · Storage: Data Storage & Governance — Permission Consumer, Thin Trust Shell ◑ Layer 1B · Retrieval: Context Management & Retrieval — Web-Scale Retrieval Primitive, No Customer Estate ○ Layer 1C · Pipelines: Data Movement & Pipelines — Agentic Movement, No Pipeline ○ Layer 2A · Orchestration: Infrastructure Orchestration — Spend-Earned Tiers, No Customer Plane ● Layer 2B · Runtime: Application Runtime & Execution — Brokered Serving, Hosted and Local Runtimes ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Product Controls, Not a Reasoning Plane ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — First-Party Value Plane: Answers, Browser, Agent **DAPM Portability Profile (components):** Retained 4 · Delegated 3 · Ceded 10 · Absent 0 **Summary:** Perplexity sells judgment, not infrastructure, and it prices that judgment per call. The row reads gap at Layer 0, 1A, 1C, 2A and 2C, moderate at 1B, and strong at 2B and Layer 3. That shape puts it in the same cohort as OpenAI and Anthropic, with one difference that matters. Those two sell intelligence and let you assemble around it. This one sells an assembled opinion about how the work should go. Presets that decide how much effort a run deserves. A classifier that decides how much searching a question needs. A ranking that decides what counts as an answer. The mechanisms underneath are largely other people's, from the frontier models it brokers to the open weights it hosts to the web itself. --- ### Qlik (Cloud Analytics + Talend Cloud + Open Lakehouse) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.9 - 4+1 v2: Visibility Read From the Cell · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/qlik.md **Decision authority (per layer):** Retained 2 · Delegated 1 · Ceded 4 · Absent 1 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Not Qlik's Layer (By Design) ◑ Layer 1A · Storage: Data Storage & Governance — Open Lakehouse on Your Substrate; Captive Trust Layer ◑ Layer 1B · Retrieval: Context Management & Retrieval — Turnkey RAG, Closed Box ● Layer 1C · Pipelines: Data Movement & Pipelines — Qlik's Heartland — CDC + Talend + Lakehouse Ingestion ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Managed Task & Cluster Orchestration; No GPU Plane ◑ Layer 2B · Runtime: Application Runtime & Execution — Turnkey Agents & Automation; No Model Serving ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Bring Your Own Reasoning Plane (MCP-First) ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Qlik's Native Layer — The Analytics Value Plane **DAPM Portability Profile (components):** Retained 0 · Delegated 1 · Ceded 19 · Absent 0 **Summary:** Qlik is a data integration and analytics software vendor that floats above Layer 0 (SaaS on AWS, and AWS-only) and is strongest at the two ends it has always owned: data movement (1C — the Attunity/Talend/Upsolver heartland) and the analytics value plane (Layer 3 — the application estate plus a GA detect-predict-act loop). The middle of the stack is uniformly moderate — storage/governance, retrieval, orchestration, and runtime are all real but product-bounded — and the reasoning plane is a gap by explicit posture. Authority sits in Qlik's opinion layer wherever Qlik ships one; the substrate never. --- ### Red Hat (OpenShift AI + RHEL AI + Ansible) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.2 - 4+1 v2: Accelerator Operators Retained · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/redhat.md **Decision authority (per layer):** Retained 2 · Delegated 6 · Ceded 0 · Absent 0 **Layer Status:** ◑ Layer 0 · Compute: Compute & Network Fabric — Hardware-Agnostic Substrate: RHEL, OpenShift Virtualization, Multi-Accelerator Enablement ◑ Layer 1A · Storage: Data Storage & Governance — Ceph-Based Storage in Platform Plus + Workload Compliance Governance; No Catalog ◑ Layer 1B · Retrieval: Context Management & Retrieval — Open-Source Retrieval Assembled on OpenShift AI ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Kafka, Debezium, Camel, and Kubeflow Pipelines on Red Hat Subscriptions; No Enterprise Lineage ● Layer 2A · Orchestration: Infrastructure Orchestration — One Control Plane for Containers, VMs, and GPUs: OpenShift, OpenShift AI Kueue, ACM, Ansible ● Layer 2B · Runtime: Application Runtime & Execution — Open Serving Stack: vLLM, llm-d, Models-as-a-Service ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Enforcing Inference Gateway; Agent Governance in Preview ◇ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Certified ISV Ecosystem; No Business Applications **DAPM Portability Profile (components):** Retained 24 · Delegated 1 · Ceded 0 · Absent 0 **Summary:** Red Hat is the instrument's open-source substrate vendor: an operating system, a Kubernetes platform, and an AI serving stack the enterprise can run on any hardware or cloud. It is strong where workloads are orchestrated (2A: OpenShift, Kueue, fleet policy) and run (2B: vLLM, llm-d, Models-as-a-Service), moderate across the data layers, a single enforcing gateway at 2C, and partner at Layer 3. Nearly every component reads Retained on the possess-and-operate line; this is the row the methodology's Red Hat standard describes. --- ### Redis (Redis Cloud + Redis Software + Redis Open Source + Redis Iris) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.5 - 4+1 v2: 1C Reconcile Note Settled · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/redis.md **Decision authority (per layer):** Retained 1 · Delegated 2 · Ceded 1 · Absent 4 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Not Redis's Layer (By Design); Redis Cloud on Redis's Cloud Accounts or Yours, Redis Software on Your Metal ◑ Layer 1A · Storage: Data Storage & Governance — In-Memory System of Record for Hot State, Geo-Distributed and Tiered, With an ML Feature Store; No Data-Estate Catalog ◑ Layer 1B · Retrieval: Context Management & Retrieval — Native Hybrid Retrieval in the Open Engine, Without GA Models; Semantic Caching, Vector Sets, Context Retriever, and Agent Memory in Preview ◑ Layer 1C · Pipelines: Data Movement & Pipelines — RDI: CDC Into Redis, Self-Hosted From Eleven Sources or Managed on Redis Cloud From AWS Sources; One Destination ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Cluster Management, Kubernetes Operator, Managed Sizing, and Flex Tuning; No GPU Plane ○ Layer 2B · Runtime: Application Runtime & Execution — A Tool Surface (MCP Server) and Database Logic (Lua Functions), Not a Runtime; Semantic Caching in Preview ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — ACLs and Scoped Keys, Not a Reasoning Plane; Context Retriever's Governed Tools in Preview ○ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Developer Tooling With a Beta Assistant; No Value Plane **DAPM Portability Profile (components):** Retained 5 · Delegated 2 · Ceded 10 · Absent 0 **Summary:** Redis is the hot state under the agent, and the map reads it as a data platform whose AI story is one layer deep in production and three layers deep in preview. Layer 1B is moderate on hybrid retrieval in the open-source engine (the Redis Query Engine with FT.HYBRID, generally available in Redis 8.4 and on Redis Cloud since March 2026) with server-side relevance stopping at fusion and the context-management pieces (semantic caching, the context retriever, agent memory) in preview; 1A, 1C, and 2A are moderate on an in-memory system of record with Active-Active geo-distribution, Flex tiering, and a generally available ML feature store, on Redis Data Integration's change data capture (CDC) into Redis, and on cluster management with a Kubernetes operator; four layers are gaps: no substrate, no model serving or agent runtime (the Model Context Protocol server and Lua Functions are a tool surface and database logic, not a runtime), no reasoning plane, and no application beyond a client whose assistant is Beta. --- ### Salesforce (Agentforce 360 + Data 360 + Customer 360) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.10 - 4+1 v2: Layer 3 Read Like Any Layer · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/salesforce.md **Decision authority (per layer):** Retained 3 · Delegated 1 · Ceded 4 · Absent 0 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Not Salesforce's Layer (By Design) ● Layer 1A · Storage: Data Storage & Governance — Authorization Authority, Dual Capture ● Layer 1B · Retrieval: Context Management & Retrieval — Governed Retrieval — Native Infrastructure Surface ● Layer 1C · Pipelines: Data Movement & Pipelines — MuleSoft + Informatica — Heterogeneous Movement ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Managed Orchestration; No GPU Plane ● Layer 2B · Runtime: Application Runtime & Execution — Atlas Runtime — Governed, Constructible, Any-Model ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Governance Without a Reasoning Mechanism ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Salesforce's Native Layer — First-Party Value Plane **DAPM Portability Profile (components):** Retained 0 · Delegated 3 · Ceded 30 · Absent 0 **Summary:** Salesforce is an enterprise application, agent, and data platform that floats above Layer 0 and is strong everywhere it has always lived: the data plane of storage and governance, retrieval, and movement, plus the agent runtime and the value plane. It moderates at the two layers it does not truly own. Infrastructure orchestration, where it manages its own workloads on rented cloud. And the reasoning plane, where it governs and enforces but does not reason. The agent runtime it does own is strong but captive-by-design: a complete runtime whose agents can never leave Salesforce. Authority sits in Salesforce's opinion layer wherever Salesforce ships one, which is nearly everywhere above the silicon. --- ### SAP Business AI Platform (Joule + Business Data Cloud + BTP) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.3 - 4+1 v2: Layer 3 Read Like Any Layer · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/sap.md **Decision authority (per layer):** Retained 3 · Delegated 4 · Ceded 1 · Absent 0 **Layer Status:** ◑ Layer 0 · Compute: Compute & Network Fabric — SAP-Operated IaaS in SAP Data Centers; Sovereign Cloud On-Site; Hyperscalers for the Rest ● Layer 1A · Storage: Data Storage & Governance — Governed Business Data Cloud: HANA, Data Products, Mastering, Zero-Copy ● Layer 1B · Retrieval: Context Management & Retrieval — In-Database Vector + Knowledge Graph + Grounding Service; Embedding Choice ◑ Layer 1C · Pipelines: Data Movement & Pipelines — iPaaS + Datasphere Replication with SAP-Source CDC + Zero-Copy Sharing; Transformation Stays Local ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Managed Kubernetes (Kyma) + AI Core GPU Instance Types + Cloud Foundry; No Fair-Share ● Layer 2B · Runtime: Application Runtime & Execution — Generative AI Hub Serving + AI Core Custom Models + Joule Studio Agents; Any-Model ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Registry GA + MCP Gateway GA + Joule Orchestration; Identity, Agent Observability, and A2A Dated H2 2026 ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — First-Party Business Applications with Joule Across the Suite; Autonomous Suite Arriving **DAPM Portability Profile (components):** Retained 0 · Delegated 7 · Ceded 26 · Absent 0 **Summary:** SAP is the largest first-party value plane on the instrument with a governed data plane underneath it and a serving-and-agents runtime between, and the row reads that way: strong at the data foundation (Layer 1A), retrieval (1B), the runtime (2B), and the applications (Layer 3), moderate at the substrate SAP now operates itself (Layer 0), at orchestration (2A), at pipelines (1C), and at the reasoning plane (2C). Authority is the SaaS shape with more seams than most: nearly everything is SAP intellectual property (IP) and Ceded, with Delegated readings where the consumed interface is a multi-vendor standard (Delta Sharing at 1A and 1C, the Kubernetes API and Cloud Foundry at 2A) or where SAP brokers models the enterprise can swap or hosts models the enterprise owns (the generative AI hub and AI Core at 2B), and one Retained decision-authority reading at 1C, where SAP's engines execute flows the enterprise wrote without judgment of their own. --- ### ServiceNow AI Platform (Now Assist + AI Control Tower + Workflow Data Fabric) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.6 - 4+1 v2: Supervised Mode Is a Gate · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/servicenow.md **Decision authority (per layer):** Retained 4 · Delegated 2 · Ceded 2 · Absent 0 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Not ServiceNow's Layer (By Design) ● Layer 1A · Storage: Data Storage & Governance — Workflow System of Record + Zero-Copy Reach + Data Catalog ◑ Layer 1B · Retrieval: Context Management & Retrieval — Platform-Scoped Hybrid Retrieval with Embedding Choice; No Infrastructure Surface ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Integration Fabric + Lineage + Cost-Aware Archive; No Enterprise ETL, No CDC ○ Layer 2A · Orchestration: Infrastructure Orchestration — Managed SaaS Lifecycle Only; No Orchestration Surface, No GPU Plane ● Layer 2B · Runtime: Application Runtime & Execution — Constructible Governed Agent Runtime; Multi-Model; Flow-Executed ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Cross-Enterprise Agent Registry, Containment, Orchestration, Observability; Identity and Gateway Legs Contested ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — First-Party Workflow Applications with an Autonomous Workforce Arriving **DAPM Portability Profile (components):** Retained 0 · Delegated 5 · Ceded 26 · Absent 0 **Summary:** ServiceNow is a workflow platform that built an agent runtime and then a governance plane for everyone else's agents, and the row reads that way: strong where agents are built and run (Layer 2B), where the buyer's work gets done (Layer 3), and at the data foundation (1A), moderate through retrieval, pipelines, and the reasoning plane, and a gap at the two infrastructure layers by design. Authority is the software-as-a-service (SaaS) shape throughout: nothing at Layer 0 or 2A is the customer's, and nearly everything scored above is ServiceNow intellectual property (IP) and Ceded, with two Delegated seams, zero-copy federation and the customer's own model keys, and one Retained reading at 1C, where the platform executes integrations the enterprise wrote without judgment of its own. --- ### Snowflake AI Data Cloud *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.7 - 4+1 v2: 2A Authority on Resize and Assignment · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/snowflake.md **Decision authority (per layer):** Retained 0 · Delegated 7 · Ceded 1 · Absent 0 **Layer Status:** ○ Layer 0 · Compute: Compute & Network Fabric — Not Snowflake's Layer (By Design) ● Layer 1A · Storage: Data Storage & Governance — Snowflake Strength: Governed Lakehouse Across Three Clouds ● Layer 1B · Retrieval: Context Management & Retrieval — Native Hybrid Retrieval + Governed Semantic Layer ● Layer 1C · Pipelines: Data Movement & Pipelines — Openflow + Streaming + Declarative Pipelines ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Managed Compute + GA GPU Pools; Platform-Scoped ● Layer 2B · Runtime: Application Runtime & Execution — Cortex Runtime: Any-Model Serving + GA Agents + ML Serving ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Deep Agent Identity, No Gateway, No Native Orchestration ● Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — First-Party Agents, Observability, and Marketplace **DAPM Portability Profile (components):** Retained 2 · Delegated 9 · Ceded 25 · Absent 0 **Summary:** Snowflake is a data and AI platform that floats above Layer 0, owning no silicon and renting all three clouds, and is strong across the data plane (1A, 1B, 1C), the application runtime (2B), and the value plane (3). It moderates only where it doesn't own the layer: infrastructure orchestration (2A), where compute pools are proprietary node-level scheduling on rented instances, and the reasoning plane (2C), where agent identity and observability are GA but the request-time gateway and native multi-agent orchestration aren't. Authority sits in Snowflake's opinion layer wherever Snowflake ships one; the substrate never. --- ### Supermicro DCBBS & AI Data Platforms *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.11 - 4+1 v2: Visibility Settled on Docs · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/supermicro.md **Decision authority (per layer):** Retained 2 · Delegated 2 · Ceded 3 · Absent 1 **Layer Status:** ● Layer 0 · Compute: Compute & Network Fabric — Supermicro Strength ● Layer 1A · Storage: Data Storage & Governance — Partner-Delivered Strength ● Layer 1B · Retrieval: Context Management & Retrieval — Partner-Delivered Retrieval ● Layer 1C · Pipelines: Data Movement & Pipelines — Partner-Delivered Pipeline Engines ● Layer 2A · Orchestration: Infrastructure Orchestration — First-Party GPU Cloud Orchestration ◑ Layer 2B · Runtime: Application Runtime & Execution — NVIDIA Passthrough + Open Substrate ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Enterprise Responsibility ◇ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Uncurated Partner Ecosystem **DAPM Portability Profile (components):** Retained 4 · Delegated 7 · Ceded 14 · Absent 0 **Summary:** Supermicro grades as the most capability-dense OEM row on this instrument — strong from Layer 0 through 2A — while owning less software IP than any peer. The mechanism is the row's defining feature, and the reason for every strong grade: Supermicro's paper productizes best-of-breed partner platforms (VAST, WEKA, DDN, Cloudian, IBM, Nutanix, Everpure) as named Supermicro products with documented single-vendor support, on top of the industry's fastest NVIDIA-generation cadence, DLC-2 liquid cooling at gigawatt scale, and a DCBBS motion that delivers whole data centers as building blocks. The capability axis credits what a customer can deploy on the vendor's paper today — the same rule that scores every OEM strong at Layer 0 on NVIDIA silicon — and by that rule Supermicro's deployable menu is the widest and strongest in the on-prem market. --- ### SUSE (Rancher Prime + SUSE AI + SUSE AI Factory with NVIDIA + SLES 16) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.5 - 4+1 v2: GPU Operator Retained · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/suse.md **Decision authority (per layer):** Retained 1 · Delegated 6 · Ceded 1 · Absent 0 **Layer Status:** ◑ Layer 0 · Compute: Compute & Network Fabric — Hardware-Agnostic Abstraction: SLES 16, SUSE Virtualization with MIG, Any Vendor's Metal ◑ Layer 1A · Storage: Data Storage & Governance — Platform Storage (Longhorn) and S3-Compatible Objects (MinIO), Not a Data Foundation ◑ Layer 1B · Retrieval: Context Management & Retrieval — Supported Open-Source Vector Stores and RAG in the AI Library; NVIDIA's RAG Blueprint Through the Factory ◑ Layer 1C · Pipelines: Data Movement & Pipelines — NVIDIA's Fixed RAG Ingest Pipeline Through the Factory; No General Pipeline Layer ● Layer 2A · Orchestration: Infrastructure Orchestration — Rancher Heritage Strength: Multi-Cluster Kubernetes, AI-Conformant RKE2, Virtual Clusters with GPU Quotas, an Apache 2.0 AI Factory Operator ◑ Layer 2B · Runtime: Application Runtime & Execution — Supported Open Inference (vLLM, Ollama) + NVIDIA AI Enterprise (NIM, NeMo, Nemotron) Through the Factory; No SUSE Runtime ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Observability for AI Workloads + Rancher AI's Supervisor, Agent Configs, and Human-Validation Gates; MCP Gateway in Preview ◑ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Liz and the Crew for Infrastructure Operations, SUSE Industrial Edge for the Plant Floor; Platform-Enabled Beyond **DAPM Portability Profile (components):** Retained 14 · Delegated 2 · Ceded 12 · Absent 0 **Summary:** SUSE is an open-source infrastructure vendor that turned its Kubernetes into an AI platform by curating what runs on it, and the map reads it strongest where Rancher always was. Layer 2A is strong on multi-cluster Kubernetes with AI-conformant RKE2, virtual clusters that hand out GPU slices under quotas, and an Apache 2.0 AI Factory operator; six layers are moderate (a hardware-agnostic abstraction with SUSE Linux Enterprise Server (SLES) 16 and Multi-Instance GPU (MIG) capable virtualization at Layer 0; Longhorn and MinIO as platform storage at 1A; supported open-source vector stores plus NVIDIA's retrieval blueprint at 1B; NVIDIA's fixed retrieval-augmented generation (RAG) ingest pipeline at 1C; open inference engines plus NVIDIA AI Enterprise through the Factory at 2B; observability and Rancher AI's supervisor and human-validation gates at 2C; Liz, the Rancher Prime MCP server, and SUSE Industrial Edge at Layer 3), and none is a gap. --- ### VAST AI Operating System *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.14 - 4+1 v2: Visibility Settled on Docs · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/vast.md **Decision authority (per layer):** Retained 2 · Delegated 2 · Ceded 3 · Absent 1 **Layer Status:** ◑ Layer 0 · Compute: Compute & Network Fabric — Software-Defined ● Layer 1A · Storage: Data Storage & Governance — VAST Strength ● Layer 1B · Retrieval: Context Management & Retrieval — VAST Strength ● Layer 1C · Pipelines: Data Movement & Pipelines — VAST Strength ◑ Layer 2A · Orchestration: Infrastructure Orchestration — Partial ● Layer 2B · Runtime: Application Runtime & Execution — VAST Strength — AgentEngine (GA) ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Gap — placement abstraction, not a reasoning plane ◇ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — Focused Ecosystem **DAPM Portability Profile (components):** Retained 0 · Delegated 5 · Ceded 29 · Absent 0 **Summary:** VAST is the most architecturally distinct vendor in this assessment series because it deliberately collapses traditionally separate infrastructure layers into a single platform. The VAST AI Operating System unifies storage (DataStore), metadata and vector database (DataBase + Catalog), global namespace (DataSpace), serverless compute (DataEngine), retrieval (InsightEngine), agent runtime (AgentEngine), governance (PolicyEngine), and model lifecycle (TuningEngine) under one authority boundary. Where Dell assembles 5–7 partner technologies to span Layers 1A through 2B, VAST provides a vertically integrated alternative with no inter-layer seams. --- ### VMware Private AI Cloud *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.15 - 4+1 v2: NVIDIA One Path of Three · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/vmware.md **Decision authority (per layer):** Retained 1 · Delegated 5 · Ceded 1 · Absent 1 **Layer Status:** ◑ Layer 0 · Compute: Compute & Network Fabric — Hardware-Agnostic Abstraction ◑ Layer 1A · Storage: Data Storage & Governance — Platform Storage, Not AI-Native ◑ Layer 1B · Retrieval: Context Management & Retrieval — Foundational RAG Services ○ Layer 1C · Pipelines: Data Movement & Pipelines — No Pipeline Layer ● Layer 2A · Orchestration: Infrastructure Orchestration — VMware Heritage Strength ◑ Layer 2B · Runtime: Application Runtime & Execution — Platform-Native, Widening Beyond NVIDIA ◑ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Agent Governance Ships, Placement Does Not ◇ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — ISV Ecosystem on a Private Cloud Platform **DAPM Portability Profile (components):** Retained 2 · Delegated 11 · Ceded 19 · Absent 0 **Summary:** VMware Private AI Cloud occupies a structurally unique position in this assessment series: it is neither an infrastructure OEM (Dell, HPE), a hyperscaler (AWS, Google Cloud), nor a data platform vendor (VAST). It is a virtualization and private cloud platform, the abstraction layer that sits between physical infrastructure and workloads. Broadcom’s strategic thesis is that VCF is ‘the permanent abstraction layer between AI software and physical chips,’ and Private AI Cloud extends that thesis into inference and agentic workloads. VMware AI Factory is its software-defined foundation, and Tanzu Platform is its agent platform. --- ### WEKA (NeuralMesh + NeuralMesh Axon + WEKApod + Augmented Memory Grid + the WEKA Operator, CSI Plugin, App Store, and AI Data Platform) *Mapped to the 4+1 Layer AI Infrastructure Model* Version: v1.3 - 4+1 v2: NetApp Calibration Updated · October 5, 2026 Sources and audit trail: https://layer2c.com/assessment/weka.md **Decision authority (per layer):** Retained 0 · Delegated 2 · Ceded 4 · Absent 2 **Layer Status:** ◑ Layer 0 · Compute: Compute & Network Fabric — NeuralMesh, Software-Defined Storage and Memory on Commodity NVMe Servers You Own, Your Cloud Accounts (AWS, Azure, GCP, OCI), or WEKApod Appliances; No Compute, No Fabric; WEKApod 3 Ships Fall 2026 ◑ Layer 1A · Storage: Data Storage & Governance — One Parallel Filesystem Behind POSIX, NFS, SMB, and a Native S3 With Object Lock, Versioning, and Audit; a Data Catalog That Indexes and Queries Filesystem Metadata at Scale; Organizations and Composable Clusters for Tenants; Encryption Under Your KMS; No Classification, No Lineage ◑ Layer 1B · Retrieval: Context Management & Retrieval — The WEKA AI Data Platform From the App Store: a Filesystem-Watching Ingest, NVIDIA NeMo Retriever Embedding, and Milvus Vector Index With POSIX Permissions Enforced at Query Time; the Engine Is NVIDIA's and Milvus, the Packaging Is WEKA's ◑ Layer 1C · Pipelines: Data Movement & Pipelines — Object-Store Tiering and Snap-to-Object for Cost-Aware Movement, AIDP's Continuous Ingest as a Fixed Pipeline; the Augmented Memory Grid's KV Cache Tiering Is Commercially Available but Undocumented; NeuralMesh 6 Replication Pending ◑ Layer 2A · Orchestration: Infrastructure Orchestration — The WEKA Operator and Cloud Deployment Manager Run NeuralMesh on the Enterprise's Kubernetes and Clouds (Custom Resources, Rolling Upgrades, Auto-Scaling Groups, Composable Tenants); a Standard CSI Door; Node Selectors and Tolerations Pin Where WEKA Runs; No Compute or GPU Scheduling ◑ Layer 2B · Runtime: Application Runtime & Execution — The AI Data Platform's Agentic RAG API: NVIDIA NIM Inference Services, a RAG Server, and WEKA's RAG Gateway and RAG Bridge Answering Over the Platform's Own Index; No General Serving, No Agent Runtime; the Augmented Memory Grid Feeds Other Engines ○ Layer 2C · Reasoning: Agentic Infrastructure — The Reasoning Plane — Tenant Isolation and Permission-Aware Retrieval Govern Data, Not Agents; Keycloak Inside AIDP Identifies Users; Not a Plane ○ Layer 3 (+1) · Applications: AI Application Layer — The Value Plane — No Application: the App Store Deploys One Documented Blueprint, the AI Data Platform, Which Is a Data Pipeline; the Catalogue's Breadth Is Undocumented **DAPM Portability Profile (components):** Retained 1 · Delegated 3 · Ceded 11 · Absent 0 **Summary:** WEKA is a storage software company that has renamed its product for the inference era: NeuralMesh (documentation version 5.1) is the WEKA Data Platform, and around it sit NeuralMesh Axon (the filesystem converged onto GPU servers), WEKApod appliances, the Augmented Memory Grid (key-value cache tiering to NVMe for inference engines), and, in 5.1, a Data Catalog, a WEKA App Store, and an AI Data Platform blueprint. The map reads it moderate at five layers and gap at three, with no strong cell. Layer 1A is the deepest cell, moderate at the top of the band: one multi-protocol namespace with a native S3 service (object lock, versioning, audit), a Data Catalog that indexes and queries filesystem metadata at scale, tenants two ways (Organizations and hardware-isolated Composable Clusters), and encryption under the enterprise's key management system (KMS), with no classification or lineage. Layer 0 is moderate on software-defined storage over servers the enterprise picks from WEKA's support matrix or clouds it rents. Layer 1B is moderate on the AI Data Platform: a filesystem watcher feeding NVIDIA NeMo Retriever and Milvus, with file permissions enforced at query time when identity is federated. Layer 1C is moderate on object-store tiering (with per-path overrides) and snap-to-object plus the platform's ingest pipeline; the Augmented Memory Grid, commercially available since November 2025, has no documentation page and is named, not scored, under rule 3: available now is not GA. Layer 2A is moderate on the WEKA Operator with node-selector pins, the Cloud Deployment Manager, Terraform modules, and an open CSI driver, with no compute scheduling. Layer 2B is moderate, with Everpure's, on the AI Data Platform's agentic RAG surface: NVIDIA's blueprint behind WEKA's gateway. Layer 2C is a gap: the platform's identity, gateway, and traces are application-scoped legs. Layer 3 is a gap: the App Store's documented catalogue is one data pipeline. --- ## Lab Validations Layer2C Labs (https://labs.layer2c.com) puts slices of the stack on real hardware and rules on where authority actually holds. Where a lab has validated a vendor's layer, it is listed below with the layers it tested. Labs validate the assessment against reality; they are not paid certifications, and the grades frozen at lab time live on each lab page. Machine corpus: https://labs.layer2c.com/labs/index.json. | Vendor | Lab | Layers validated | Lab URL | |---|---|---|---| | AMD Instinct, EPYC & Enterprise AI Suite | Lab 017: Beyond CUDA: the lock was never the silicon | Layer 0, Layer 2B, Layer 2C | https://labs.layer2c.com/labs/beyond-cuda | | AWS AI Infrastructure | Lab 012: Buy the harness, not the tier | Layer 0 | https://labs.layer2c.com/labs/harness-or-tier | | AWS AI Infrastructure | Lab 009: Recovered capacity is real, and it fails honest | Layer 0 | https://labs.layer2c.com/labs/xeon-rematch | | AWS AI Infrastructure | Lab 002: Own the weights, or the platform owns you | Layer 2B, Layer 2C | https://labs.layer2c.com/labs/fine-tune-economics | | AWS AI Infrastructure | Lab 001: Borrow the vendor’s plumbing, not its judgment | Layer 1A, Layer 1B | https://labs.layer2c.com/labs/spark-s3vectors | | Google Cloud AI Infrastructure | Lab 012: Buy the harness, not the tier | Layer 0, Layer 2C | https://labs.layer2c.com/labs/harness-or-tier | | Google Cloud AI Infrastructure | Lab 008: Renting the chip was the easy part | Layer 0 | https://labs.layer2c.com/labs/gemma4-tpu-inference | | Google Cloud AI Infrastructure | Lab 007: The CPU exit is a batch lane, not a serving lane | Layer 0 | https://labs.layer2c.com/labs/gemma4-xeon-inference | | Google Cloud AI Infrastructure | Lab 006: Put the judgment in the constraints, not the weights | Layer 2B | https://labs.layer2c.com/labs/constraints-not-weights | | Google Cloud AI Infrastructure | Lab 004: You can’t automate a process you haven’t encoded | Layer 1A, Layer 1B, Layer 2B, Layer 2C | https://labs.layer2c.com/labs/migration-control-plane | | Intel AI Infrastructure Portfolio | Lab 009: Recovered capacity is real, and it fails honest | Layer 0, Layer 2B | https://labs.layer2c.com/labs/xeon-rematch | | Mistral AI | Lab 013: The floor is trained, not sized | Layer 2C | https://labs.layer2c.com/labs/self-host-floor | | NVIDIA AI Platform | Lab 017: Beyond CUDA: the lock was never the silicon | Layer 0, Layer 2B | https://labs.layer2c.com/labs/beyond-cuda | | NVIDIA AI Platform | Lab 013: The floor is trained, not sized | Layer 0 | https://labs.layer2c.com/labs/self-host-floor | | NVIDIA AI Platform | Lab 012: Buy the harness, not the tier | Layer 0, Layer 2C | https://labs.layer2c.com/labs/harness-or-tier | | NVIDIA AI Platform | Lab 011: Buy the labor, not the judgment | Layer 0, Layer 2C | https://labs.layer2c.com/labs/labor-not-judgment | | NVIDIA AI Platform | Lab 009: Recovered capacity is real, and it fails honest | Layer 0, Layer 2B, Layer 2C | https://labs.layer2c.com/labs/xeon-rematch | | NVIDIA AI Platform | Lab 006: Put the judgment in the constraints, not the weights | Layer 0, Layer 2B | https://labs.layer2c.com/labs/constraints-not-weights | | NVIDIA AI Platform | Lab 004: You can’t automate a process you haven’t encoded | Layer 0, Layer 2C | https://labs.layer2c.com/labs/migration-control-plane | | NVIDIA AI Platform | Lab 003: The validator determines done, not the loop | Layer 0, Layer 2C | https://labs.layer2c.com/labs/loop-control | | NVIDIA AI Platform | Lab 002: Own the weights, or the platform owns you | Layer 0, Layer 2B, Layer 2C | https://labs.layer2c.com/labs/fine-tune-economics | | NVIDIA AI Platform | Lab 001: Borrow the vendor’s plumbing, not its judgment | Layer 0, Layer 2B, Layer 2C | https://labs.layer2c.com/labs/spark-s3vectors | ## Governed chat Governed chat (vCTOA): ask the research system at any property; every answer carries typed evidence, versions, and limits. Machine door + policy: https://virtual-cto-api-1006606217023.us-east4.run.app/api/v1/vctoa/registry ## Site Navigation | Page | URL | Description | |---|---|---| | Home / Vendor Catalog | https://layer2c.com/ | Vendor catalog: filterable DAPM profiles (six per page) plus a full directory table of every assessed vendor | | Cross-Vendor Compare | https://layer2c.com/compare | Layer status matrix and DAPM heat map across all vendors | | Head-to-Head Compare | https://layer2c.com/compare/{vendorA}/{vendorB} | Any two vendors side by side, layer by layer: capability grade, decision authority, and component portability read unchanged from the published rows. Uses the vendor ids from the assessment URLs; reports differences without interpreting them | | What Is Layer 2C? | https://layer2c.com/what-is-layer-2c | Framework explainer, 4+1 model, DAPM definitions | | Methodology | https://layer2c.com/methodology | How assessments are scored and what they evaluate | | System Map | https://layer2c.com/system | How the four Layer2C properties fit as one research system | | Design Challenge | https://layer2c.com/design-challenge | Four-act reference experience: conversation, challenged architecture, measured lab, and a published record the field can cite | | AMD Instinct, EPYC & Enterprise AI Suite Assessment | https://layer2c.com/assessment/amd | Mapped to the 4+1 Layer AI Infrastructure Model | | Anthropic (Claude Platform + Claude Enterprise + Claude Code) Assessment | https://layer2c.com/assessment/anthropic | Mapped to the 4+1 Layer AI Infrastructure Model | | Articul8 AI Platform Assessment | https://layer2c.com/assessment/articul8 | Mapped to the 4+1 Layer AI Infrastructure Model | | AWS AI Infrastructure Assessment | https://layer2c.com/assessment/aws | Mapped to the 4+1 Layer AI Infrastructure Model | | Microsoft Azure AI Infrastructure Assessment | https://layer2c.com/assessment/azure | Mapped to the 4+1 Layer AI Infrastructure Model | | Cerebras Systems (CS-3 and CS-4 Systems + Cerebras Cloud Inference and Training + the Model Zoo) Assessment | https://layer2c.com/assessment/cerebras | Mapped to the 4+1 Layer AI Infrastructure Model | | Cisco Secure AI Factory with NVIDIA Assessment | https://layer2c.com/assessment/cisco | Mapped to the 4+1 Layer AI Infrastructure Model | | Cloudera (Anywhere Cloud + Data Services + Cloudera AI) Assessment | https://layer2c.com/assessment/cloudera | Mapped to the 4+1 Layer AI Infrastructure Model | | Cloudflare (Workers AI + AI Gateway + Agents SDK + Developer Platform) Assessment | https://layer2c.com/assessment/cloudflare | Mapped to the 4+1 Layer AI Infrastructure Model | | Cohere (Command + Embed + Rerank + North) Assessment | https://layer2c.com/assessment/cohere | Mapped to the 4+1 Layer AI Infrastructure Model | | Confluent, an IBM company (Confluent Cloud + Confluent Platform + Confluent Private Cloud + WarpStream + Confluent Intelligence) Assessment | https://layer2c.com/assessment/confluent | Mapped to the 4+1 Layer AI Infrastructure Model | | CoreWeave AI Cloud Assessment | https://layer2c.com/assessment/coreweave | Mapped to the 4+1 Layer AI Infrastructure Model | | Databricks Data Intelligence Platform Assessment | https://layer2c.com/assessment/databricks | Mapped to the 4+1 Layer AI Infrastructure Model | | Dataiku (DSS 15 + LLM Mesh + Agent Hub + Govern + Dataiku Cloud) Assessment | https://layer2c.com/assessment/dataiku | Mapped to the 4+1 Layer AI Infrastructure Model | | Dell AI Factory with NVIDIA Assessment | https://layer2c.com/assessment/dell | Mapped to the 4+1 Layer AI Infrastructure Model | | Elastic (Elasticsearch + Agent Builder + Security + Observability) Assessment | https://layer2c.com/assessment/elastic | Mapped to the 4+1 Layer AI Infrastructure Model | | Everpure, formerly Pure Storage (Everpure Platform: FlashArray + FlashBlade + Purity + Fusion + Pure1 + Evergreen + Portworx + Data Intelligence + Data Stream) Assessment | https://layer2c.com/assessment/everpure | Mapped to the 4+1 Layer AI Infrastructure Model | | Google Cloud AI Infrastructure Assessment | https://layer2c.com/assessment/gcp | Mapped to the 4+1 Layer AI Infrastructure Model | | Groq (GroqCloud on the LPU + Compound + Server-Side Web Retrieval; Independent Under a Non-Exclusive NVIDIA Technology License) Assessment | https://layer2c.com/assessment/groq | Mapped to the 4+1 Layer AI Infrastructure Model | | HPE Private Cloud AI Assessment | https://layer2c.com/assessment/hpe | Mapped to the 4+1 Layer AI Infrastructure Model | | Hugging Face (Hub + Team and Enterprise Plans + Storage Buckets + Inference Endpoints + Inference Providers + Jobs + Spaces + the Open-Source Libraries) Assessment | https://layer2c.com/assessment/huggingface | Mapped to the 4+1 Layer AI Infrastructure Model | | IBM Cloud Assessment | https://layer2c.com/assessment/ibm-cloud | Mapped to the 4+1 Layer AI Infrastructure Model | | IBM (watsonx + IBM Storage + Confluent + Power/Z) Assessment | https://layer2c.com/assessment/ibm | Mapped to the 4+1 Layer AI Infrastructure Model | | Intel AI Infrastructure Portfolio Assessment | https://layer2c.com/assessment/intel | Mapped to the 4+1 Layer AI Infrastructure Model | | Kamiwaza AI Orchestration Platform Assessment | https://layer2c.com/assessment/kamiwaza | Mapped to the 4+1 Layer AI Infrastructure Model | | Lenovo Hybrid AI Advantage with NVIDIA Assessment | https://layer2c.com/assessment/lenovo | Mapped to the 4+1 Layer AI Infrastructure Model | | Mistral AI Assessment | https://layer2c.com/assessment/mistral | Mapped to the 4+1 Layer AI Infrastructure Model | | MongoDB (Atlas + Enterprise Advanced + Community + Voyage AI) Assessment | https://layer2c.com/assessment/mongodb | Mapped to the 4+1 Layer AI Infrastructure Model | | Nebius (Nebius AI Cloud: GPU Compute, Managed Kubernetes and Soperator, Storage, Virtual Networks + Token Factory Inference and Fine-Tuning + the Applications Marketplace) Assessment | https://layer2c.com/assessment/nebius | Mapped to the 4+1 Layer AI Infrastructure Model | | NetApp Intelligent Data Infrastructure Assessment | https://layer2c.com/assessment/netapp | Mapped to the 4+1 Layer AI Infrastructure Model | | Nutanix Cloud Platform with Enterprise AI Assessment | https://layer2c.com/assessment/nutanix | Mapped to the 4+1 Layer AI Infrastructure Model | | NVIDIA AI Platform Assessment | https://layer2c.com/assessment/nvidia | Mapped to the 4+1 Layer AI Infrastructure Model | | Oracle Cloud Infrastructure (OCI) AI Infrastructure Assessment | https://layer2c.com/assessment/oci | Mapped to the 4+1 Layer AI Infrastructure Model | | OpenAI (API Platform + ChatGPT Enterprise + Codex) Assessment | https://layer2c.com/assessment/openai | Mapped to the 4+1 Layer AI Infrastructure Model | | Palantir AIP + Foundry + Apollo Assessment | https://layer2c.com/assessment/palantir | Mapped to the 4+1 Layer AI Infrastructure Model | | Perplexity (API Platform + Enterprise + Comet + Computer) Assessment | https://layer2c.com/assessment/perplexity | Mapped to the 4+1 Layer AI Infrastructure Model | | Qlik (Cloud Analytics + Talend Cloud + Open Lakehouse) Assessment | https://layer2c.com/assessment/qlik | Mapped to the 4+1 Layer AI Infrastructure Model | | Red Hat (OpenShift AI + RHEL AI + Ansible) Assessment | https://layer2c.com/assessment/redhat | Mapped to the 4+1 Layer AI Infrastructure Model | | Redis (Redis Cloud + Redis Software + Redis Open Source + Redis Iris) Assessment | https://layer2c.com/assessment/redis | Mapped to the 4+1 Layer AI Infrastructure Model | | Salesforce (Agentforce 360 + Data 360 + Customer 360) Assessment | https://layer2c.com/assessment/salesforce | Mapped to the 4+1 Layer AI Infrastructure Model | | SAP Business AI Platform (Joule + Business Data Cloud + BTP) Assessment | https://layer2c.com/assessment/sap | Mapped to the 4+1 Layer AI Infrastructure Model | | ServiceNow AI Platform (Now Assist + AI Control Tower + Workflow Data Fabric) Assessment | https://layer2c.com/assessment/servicenow | Mapped to the 4+1 Layer AI Infrastructure Model | | Snowflake AI Data Cloud Assessment | https://layer2c.com/assessment/snowflake | Mapped to the 4+1 Layer AI Infrastructure Model | | Supermicro DCBBS & AI Data Platforms Assessment | https://layer2c.com/assessment/supermicro | Mapped to the 4+1 Layer AI Infrastructure Model | | SUSE (Rancher Prime + SUSE AI + SUSE AI Factory with NVIDIA + SLES 16) Assessment | https://layer2c.com/assessment/suse | Mapped to the 4+1 Layer AI Infrastructure Model | | VAST AI Operating System Assessment | https://layer2c.com/assessment/vast | Mapped to the 4+1 Layer AI Infrastructure Model | | VMware Private AI Cloud Assessment | https://layer2c.com/assessment/vmware | Mapped to the 4+1 Layer AI Infrastructure Model | | WEKA (NeuralMesh + NeuralMesh Axon + WEKApod + Augmented Memory Grid + the WEKA Operator, CSI Plugin, App Store, and AI Data Platform) Assessment | https://layer2c.com/assessment/weka | Mapped to the 4+1 Layer AI Infrastructure Model | ## Sister Sites Part of the layer2c.com family — companion tools and assessments. | Site | URL | Description | |---|---|---| | Fourth Cloud | https://cloud.layer2c.com/ | Assesses operating-model readiness: on-prem control planes scored FC-0 through FC-4 | | Labs | https://labs.layer2c.com/ | Validates where authority actually holds, through scoped real-hardware builds. Machine corpus: https://labs.layer2c.com/labs/index.json (index) and https://labs.layer2c.com/llms.txt | | StackBuilder | https://stackbuilder.layer2c.com/ | Composes reference architectures from the assessed evidence | ## Plain-Text Mirrors | Resource | URL | Contents | |---|---|---| | llms.txt (this file) | https://layer2c.com/llms.txt | Index of all assessments with summaries | | llms-full.txt | https://layer2c.com/llms-full.txt | Complete layer-by-layer data for all vendors | | feed.xml | https://layer2c.com/feed.xml | RSS: new items on every assessment version | | home.md | https://layer2c.com/home.md | Plain-text homepage — vendor table, framework overview, navigation | | experience.md | https://layer2c.com/experience.md | How the site works from a buyer's and vendor's perspective — UX, workflow, interaction model | | Design Challenge llms.txt | https://layer2c.com/design-challenge/llms.txt | Machine-readable narrative, artifacts, results, limits, disclosures, and field prompts | | AMD Instinct, EPYC & Enterprise AI Suite (markdown) | https://layer2c.com/assessment/amd.md | Full assessment in markdown | | AMD Instinct, EPYC & Enterprise AI Suite (JSON) | https://layer2c.com/assessment/amd.json | Structured source JSON — same data, machine-readable | | Anthropic (Claude Platform + Claude Enterprise + Claude Code) (markdown) | https://layer2c.com/assessment/anthropic.md | Full assessment in markdown | | Anthropic (Claude Platform + Claude Enterprise + Claude Code) (JSON) | https://layer2c.com/assessment/anthropic.json | Structured source JSON — same data, machine-readable | | Articul8 AI Platform (markdown) | https://layer2c.com/assessment/articul8.md | Full assessment in markdown | | Articul8 AI Platform (JSON) | https://layer2c.com/assessment/articul8.json | Structured source JSON — same data, machine-readable | | AWS AI Infrastructure (markdown) | https://layer2c.com/assessment/aws.md | Full assessment in markdown | | AWS AI Infrastructure (JSON) | https://layer2c.com/assessment/aws.json | Structured source JSON — same data, machine-readable | | Microsoft Azure AI Infrastructure (markdown) | https://layer2c.com/assessment/azure.md | Full assessment in markdown | | Microsoft Azure AI Infrastructure (JSON) | https://layer2c.com/assessment/azure.json | Structured source JSON — same data, machine-readable | | Cerebras Systems (CS-3 and CS-4 Systems + Cerebras Cloud Inference and Training + the Model Zoo) (markdown) | https://layer2c.com/assessment/cerebras.md | Full assessment in markdown | | Cerebras Systems (CS-3 and CS-4 Systems + Cerebras Cloud Inference and Training + the Model Zoo) (JSON) | https://layer2c.com/assessment/cerebras.json | Structured source JSON — same data, machine-readable | | Cisco Secure AI Factory with NVIDIA (markdown) | https://layer2c.com/assessment/cisco.md | Full assessment in markdown | | Cisco Secure AI Factory with NVIDIA (JSON) | https://layer2c.com/assessment/cisco.json | Structured source JSON — same data, machine-readable | | Cloudera (Anywhere Cloud + Data Services + Cloudera AI) (markdown) | https://layer2c.com/assessment/cloudera.md | Full assessment in markdown | | Cloudera (Anywhere Cloud + Data Services + Cloudera AI) (JSON) | https://layer2c.com/assessment/cloudera.json | Structured source JSON — same data, machine-readable | | Cloudflare (Workers AI + AI Gateway + Agents SDK + Developer Platform) (markdown) | https://layer2c.com/assessment/cloudflare.md | Full assessment in markdown | | Cloudflare (Workers AI + AI Gateway + Agents SDK + Developer Platform) (JSON) | https://layer2c.com/assessment/cloudflare.json | Structured source JSON — same data, machine-readable | | Cohere (Command + Embed + Rerank + North) (markdown) | https://layer2c.com/assessment/cohere.md | Full assessment in markdown | | Cohere (Command + Embed + Rerank + North) (JSON) | https://layer2c.com/assessment/cohere.json | Structured source JSON — same data, machine-readable | | Confluent, an IBM company (Confluent Cloud + Confluent Platform + Confluent Private Cloud + WarpStream + Confluent Intelligence) (markdown) | https://layer2c.com/assessment/confluent.md | Full assessment in markdown | | Confluent, an IBM company (Confluent Cloud + Confluent Platform + Confluent Private Cloud + WarpStream + Confluent Intelligence) (JSON) | https://layer2c.com/assessment/confluent.json | Structured source JSON — same data, machine-readable | | CoreWeave AI Cloud (markdown) | https://layer2c.com/assessment/coreweave.md | Full assessment in markdown | | CoreWeave AI Cloud (JSON) | https://layer2c.com/assessment/coreweave.json | Structured source JSON — same data, machine-readable | | Databricks Data Intelligence Platform (markdown) | https://layer2c.com/assessment/databricks.md | Full assessment in markdown | | Databricks Data Intelligence Platform (JSON) | https://layer2c.com/assessment/databricks.json | Structured source JSON — same data, machine-readable | | Dataiku (DSS 15 + LLM Mesh + Agent Hub + Govern + Dataiku Cloud) (markdown) | https://layer2c.com/assessment/dataiku.md | Full assessment in markdown | | Dataiku (DSS 15 + LLM Mesh + Agent Hub + Govern + Dataiku Cloud) (JSON) | https://layer2c.com/assessment/dataiku.json | Structured source JSON — same data, machine-readable | | Dell AI Factory with NVIDIA (markdown) | https://layer2c.com/assessment/dell.md | Full assessment in markdown | | Dell AI Factory with NVIDIA (JSON) | https://layer2c.com/assessment/dell.json | Structured source JSON — same data, machine-readable | | Elastic (Elasticsearch + Agent Builder + Security + Observability) (markdown) | https://layer2c.com/assessment/elastic.md | Full assessment in markdown | | Elastic (Elasticsearch + Agent Builder + Security + Observability) (JSON) | https://layer2c.com/assessment/elastic.json | Structured source JSON — same data, machine-readable | | Everpure, formerly Pure Storage (Everpure Platform: FlashArray + FlashBlade + Purity + Fusion + Pure1 + Evergreen + Portworx + Data Intelligence + Data Stream) (markdown) | https://layer2c.com/assessment/everpure.md | Full assessment in markdown | | Everpure, formerly Pure Storage (Everpure Platform: FlashArray + FlashBlade + Purity + Fusion + Pure1 + Evergreen + Portworx + Data Intelligence + Data Stream) (JSON) | https://layer2c.com/assessment/everpure.json | Structured source JSON — same data, machine-readable | | Google Cloud AI Infrastructure (markdown) | https://layer2c.com/assessment/gcp.md | Full assessment in markdown | | Google Cloud AI Infrastructure (JSON) | https://layer2c.com/assessment/gcp.json | Structured source JSON — same data, machine-readable | | Groq (GroqCloud on the LPU + Compound + Server-Side Web Retrieval; Independent Under a Non-Exclusive NVIDIA Technology License) (markdown) | https://layer2c.com/assessment/groq.md | Full assessment in markdown | | Groq (GroqCloud on the LPU + Compound + Server-Side Web Retrieval; Independent Under a Non-Exclusive NVIDIA Technology License) (JSON) | https://layer2c.com/assessment/groq.json | Structured source JSON — same data, machine-readable | | HPE Private Cloud AI (markdown) | https://layer2c.com/assessment/hpe.md | Full assessment in markdown | | HPE Private Cloud AI (JSON) | https://layer2c.com/assessment/hpe.json | Structured source JSON — same data, machine-readable | | Hugging Face (Hub + Team and Enterprise Plans + Storage Buckets + Inference Endpoints + Inference Providers + Jobs + Spaces + the Open-Source Libraries) (markdown) | https://layer2c.com/assessment/huggingface.md | Full assessment in markdown | | Hugging Face (Hub + Team and Enterprise Plans + Storage Buckets + Inference Endpoints + Inference Providers + Jobs + Spaces + the Open-Source Libraries) (JSON) | https://layer2c.com/assessment/huggingface.json | Structured source JSON — same data, machine-readable | | IBM Cloud (markdown) | https://layer2c.com/assessment/ibm-cloud.md | Full assessment in markdown | | IBM Cloud (JSON) | https://layer2c.com/assessment/ibm-cloud.json | Structured source JSON — same data, machine-readable | | IBM (watsonx + IBM Storage + Confluent + Power/Z) (markdown) | https://layer2c.com/assessment/ibm.md | Full assessment in markdown | | IBM (watsonx + IBM Storage + Confluent + Power/Z) (JSON) | https://layer2c.com/assessment/ibm.json | Structured source JSON — same data, machine-readable | | Intel AI Infrastructure Portfolio (markdown) | https://layer2c.com/assessment/intel.md | Full assessment in markdown | | Intel AI Infrastructure Portfolio (JSON) | https://layer2c.com/assessment/intel.json | Structured source JSON — same data, machine-readable | | Kamiwaza AI Orchestration Platform (markdown) | https://layer2c.com/assessment/kamiwaza.md | Full assessment in markdown | | Kamiwaza AI Orchestration Platform (JSON) | https://layer2c.com/assessment/kamiwaza.json | Structured source JSON — same data, machine-readable | | Lenovo Hybrid AI Advantage with NVIDIA (markdown) | https://layer2c.com/assessment/lenovo.md | Full assessment in markdown | | Lenovo Hybrid AI Advantage with NVIDIA (JSON) | https://layer2c.com/assessment/lenovo.json | Structured source JSON — same data, machine-readable | | Mistral AI (markdown) | https://layer2c.com/assessment/mistral.md | Full assessment in markdown | | Mistral AI (JSON) | https://layer2c.com/assessment/mistral.json | Structured source JSON — same data, machine-readable | | MongoDB (Atlas + Enterprise Advanced + Community + Voyage AI) (markdown) | https://layer2c.com/assessment/mongodb.md | Full assessment in markdown | | MongoDB (Atlas + Enterprise Advanced + Community + Voyage AI) (JSON) | https://layer2c.com/assessment/mongodb.json | Structured source JSON — same data, machine-readable | | Nebius (Nebius AI Cloud: GPU Compute, Managed Kubernetes and Soperator, Storage, Virtual Networks + Token Factory Inference and Fine-Tuning + the Applications Marketplace) (markdown) | https://layer2c.com/assessment/nebius.md | Full assessment in markdown | | Nebius (Nebius AI Cloud: GPU Compute, Managed Kubernetes and Soperator, Storage, Virtual Networks + Token Factory Inference and Fine-Tuning + the Applications Marketplace) (JSON) | https://layer2c.com/assessment/nebius.json | Structured source JSON — same data, machine-readable | | NetApp Intelligent Data Infrastructure (markdown) | https://layer2c.com/assessment/netapp.md | Full assessment in markdown | | NetApp Intelligent Data Infrastructure (JSON) | https://layer2c.com/assessment/netapp.json | Structured source JSON — same data, machine-readable | | Nutanix Cloud Platform with Enterprise AI (markdown) | https://layer2c.com/assessment/nutanix.md | Full assessment in markdown | | Nutanix Cloud Platform with Enterprise AI (JSON) | https://layer2c.com/assessment/nutanix.json | Structured source JSON — same data, machine-readable | | NVIDIA AI Platform (markdown) | https://layer2c.com/assessment/nvidia.md | Full assessment in markdown | | NVIDIA AI Platform (JSON) | https://layer2c.com/assessment/nvidia.json | Structured source JSON — same data, machine-readable | | Oracle Cloud Infrastructure (OCI) AI Infrastructure (markdown) | https://layer2c.com/assessment/oci.md | Full assessment in markdown | | Oracle Cloud Infrastructure (OCI) AI Infrastructure (JSON) | https://layer2c.com/assessment/oci.json | Structured source JSON — same data, machine-readable | | OpenAI (API Platform + ChatGPT Enterprise + Codex) (markdown) | https://layer2c.com/assessment/openai.md | Full assessment in markdown | | OpenAI (API Platform + ChatGPT Enterprise + Codex) (JSON) | https://layer2c.com/assessment/openai.json | Structured source JSON — same data, machine-readable | | Palantir AIP + Foundry + Apollo (markdown) | https://layer2c.com/assessment/palantir.md | Full assessment in markdown | | Palantir AIP + Foundry + Apollo (JSON) | https://layer2c.com/assessment/palantir.json | Structured source JSON — same data, machine-readable | | Perplexity (API Platform + Enterprise + Comet + Computer) (markdown) | https://layer2c.com/assessment/perplexity.md | Full assessment in markdown | | Perplexity (API Platform + Enterprise + Comet + Computer) (JSON) | https://layer2c.com/assessment/perplexity.json | Structured source JSON — same data, machine-readable | | Qlik (Cloud Analytics + Talend Cloud + Open Lakehouse) (markdown) | https://layer2c.com/assessment/qlik.md | Full assessment in markdown | | Qlik (Cloud Analytics + Talend Cloud + Open Lakehouse) (JSON) | https://layer2c.com/assessment/qlik.json | Structured source JSON — same data, machine-readable | | Red Hat (OpenShift AI + RHEL AI + Ansible) (markdown) | https://layer2c.com/assessment/redhat.md | Full assessment in markdown | | Red Hat (OpenShift AI + RHEL AI + Ansible) (JSON) | https://layer2c.com/assessment/redhat.json | Structured source JSON — same data, machine-readable | | Redis (Redis Cloud + Redis Software + Redis Open Source + Redis Iris) (markdown) | https://layer2c.com/assessment/redis.md | Full assessment in markdown | | Redis (Redis Cloud + Redis Software + Redis Open Source + Redis Iris) (JSON) | https://layer2c.com/assessment/redis.json | Structured source JSON — same data, machine-readable | | Salesforce (Agentforce 360 + Data 360 + Customer 360) (markdown) | https://layer2c.com/assessment/salesforce.md | Full assessment in markdown | | Salesforce (Agentforce 360 + Data 360 + Customer 360) (JSON) | https://layer2c.com/assessment/salesforce.json | Structured source JSON — same data, machine-readable | | SAP Business AI Platform (Joule + Business Data Cloud + BTP) (markdown) | https://layer2c.com/assessment/sap.md | Full assessment in markdown | | SAP Business AI Platform (Joule + Business Data Cloud + BTP) (JSON) | https://layer2c.com/assessment/sap.json | Structured source JSON — same data, machine-readable | | ServiceNow AI Platform (Now Assist + AI Control Tower + Workflow Data Fabric) (markdown) | https://layer2c.com/assessment/servicenow.md | Full assessment in markdown | | ServiceNow AI Platform (Now Assist + AI Control Tower + Workflow Data Fabric) (JSON) | https://layer2c.com/assessment/servicenow.json | Structured source JSON — same data, machine-readable | | Snowflake AI Data Cloud (markdown) | https://layer2c.com/assessment/snowflake.md | Full assessment in markdown | | Snowflake AI Data Cloud (JSON) | https://layer2c.com/assessment/snowflake.json | Structured source JSON — same data, machine-readable | | Supermicro DCBBS & AI Data Platforms (markdown) | https://layer2c.com/assessment/supermicro.md | Full assessment in markdown | | Supermicro DCBBS & AI Data Platforms (JSON) | https://layer2c.com/assessment/supermicro.json | Structured source JSON — same data, machine-readable | | SUSE (Rancher Prime + SUSE AI + SUSE AI Factory with NVIDIA + SLES 16) (markdown) | https://layer2c.com/assessment/suse.md | Full assessment in markdown | | SUSE (Rancher Prime + SUSE AI + SUSE AI Factory with NVIDIA + SLES 16) (JSON) | https://layer2c.com/assessment/suse.json | Structured source JSON — same data, machine-readable | | VAST AI Operating System (markdown) | https://layer2c.com/assessment/vast.md | Full assessment in markdown | | VAST AI Operating System (JSON) | https://layer2c.com/assessment/vast.json | Structured source JSON — same data, machine-readable | | VMware Private AI Cloud (markdown) | https://layer2c.com/assessment/vmware.md | Full assessment in markdown | | VMware Private AI Cloud (JSON) | https://layer2c.com/assessment/vmware.json | Structured source JSON — same data, machine-readable | | WEKA (NeuralMesh + NeuralMesh Axon + WEKApod + Augmented Memory Grid + the WEKA Operator, CSI Plugin, App Store, and AI Data Platform) (markdown) | https://layer2c.com/assessment/weka.md | Full assessment in markdown | | WEKA (NeuralMesh + NeuralMesh Axon + WEKApod + Augmented Memory Grid + the WEKA Operator, CSI Plugin, App Store, and AI Data Platform) (JSON) | https://layer2c.com/assessment/weka.json | Structured source JSON — same data, machine-readable |