{ "edition": "2026-10-13", "years": [ 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024, 2025, 2026 ], "ai_labels": [ [ "infra", "AI workload support" ], [ "stack", "AI platform stack" ], [ "ops", "AI-assisted operations" ], [ "agents", "Agent openness" ] ], "index_weights": { "cap": 35, "cred": 20, "safety": 20, "ai": 10, "integ": 10, "mom": 5 }, "domains": [ { "key": "lifecycle", "label": "Versions, upgrades and support", "weight": 12 }, { "key": "compute", "label": "Nodes, autoscaling and compute choice", "weight": 12 }, { "key": "network", "label": "Networking", "weight": 10 }, { "key": "storage", "label": "Storage and data protection", "weight": 8 }, { "key": "security", "label": "Security and compliance", "weight": 12 }, { "key": "ops", "label": "Observability and day-2 operations", "weight": 8 }, { "key": "fleet", "label": "Multi-cluster, hybrid and fleet", "weight": 10 }, { "key": "devplat", "label": "Developer and app platform", "weight": 8 }, { "key": "accel", "label": "GPU and AI workloads", "weight": 10 }, { "key": "sovereignty", "label": "Regions, sovereignty and SLA", "weight": 10 } ], "platforms": [ { "key": "eks", "name": "Amazon Elastic Kubernetes Service (EKS), incl. Auto Mode, Fargate, Hybrid Nodes", "short": "Amazon EKS", "kind": "product", "group": "hyperscaler", "since": 2018, "archetype": "Broad, fast, paid by the hour", "grade": "A", "driver": "C", "mix": { "V": 13, "C": 40, "M": 27, "U": 7, "P": 13 }, "idx": 0.16, "domains": { "lifecycle": { "2016": null, "2017": null, "2018": 2.5, "2019": 3, "2020": 3, "2021": 3, "2022": 3, "2023": 3.5, "2024": 4, "2025": 4, "2026": 4 }, "compute": { "2016": null, "2017": null, "2018": 2.5, "2019": 3.5, "2020": 4, "2021": 4, "2022": 4, "2023": 4, "2024": 4.5, "2025": 5, "2026": 5 }, "network": { "2016": null, "2017": null, "2018": 2.5, "2019": 3, "2020": 3, "2021": 3, "2022": 3.5, "2023": 4, "2024": 4, "2025": 4, "2026": 4 }, "storage": { "2016": null, "2017": null, "2018": 2, "2019": 2, "2020": 2, "2021": 2, "2022": 3, "2023": 3.5, "2024": 3.5, "2025": 4.5, "2026": 4.5 }, "security": { "2016": null, "2017": null, "2018": 2.5, "2019": 3.5, "2020": 4, "2021": 4, "2022": 4, "2023": 4, "2024": 4, "2025": 4, "2026": 4 }, "ops": { "2016": null, "2017": null, "2018": 1.5, "2019": 2.5, "2020": 2.5, "2021": 3, "2022": 3, "2023": 3.5, "2024": 4.5, "2025": 4.5, "2026": 4.5 }, "fleet": { "2016": null, "2017": null, "2018": 1, "2019": 1.5, "2020": 1.5, "2021": 2.5, "2022": 3, "2023": 3, "2024": 3.5, "2025": 4, "2026": 4 }, "devplat": { "2016": null, "2017": null, "2018": 1.5, "2019": 1.5, "2020": 2.5, "2021": 2.5, "2022": 3, "2023": 3, "2024": 3, "2025": 4, "2026": 4 }, "accel": { "2016": null, "2017": null, "2018": 2, "2019": 2, "2020": 2, "2021": 2, "2022": 3, "2023": 3, "2024": 4, "2025": 4.5, "2026": 4.5 }, "sovereignty": { "2016": null, "2017": null, "2018": 2.5, "2019": 3, "2020": 3.5, "2021": 3.5, "2022": 4, "2023": 4, "2024": 4, "2025": 4, "2026": 4.5 } }, "cred": [ [ "Cadence", 5, "Dated document history and a kept version calendar; 1.27 to 1.37 each reached EKS 36 to 44 days after upstream" ], [ "Roadmap transparency", 5, "Public containers-roadmap board with labels and issue history since 2018" ], [ "Say-do", 3, "2 of 4 resolved promises GA within 12 months; restricted labels 'Coming Soon' since 2023-03, Fargate GPU never shipped for EKS" ], [ "Velocity", 5, "Several substantive launches a quarter: Auto Mode, Hybrid Nodes, Backup, Capabilities, Provisioned Control Plane" ], [ "Lifecycle stability", 4.5, "26 months per minor with extended support, rollback within 7 days; forced upgrade at end of extended support" ] ], "comm": [ [ "Price and licence volatility", 3, "Extended support fee from 2024-04 at six times standard; Auto Mode, Hybrid Nodes and Provisioned Control Plane fees added 2024 to 2025" ], [ "Purchase constraints", 1, "Pay as you go with no minimum, but no free control-plane tier" ], [ "Channel and access", 0.5, "Self-service sign-up, public docs and pricing, worldwide regions" ], [ "Owner stability", 0.5, "Amazon, stable and profitable" ], [ "Cost of staying supported", 3, "14 months standard; staying longer costs $0.60 per cluster-hour; control plane auto-upgraded after 26 months" ] ], "risk": "commercial", "lock": [ [ "Formats", 1.5, "Conformant Kubernetes; Auto Mode NodeClass and Karpenter CRDs are provider-specific but workloads use standard APIs" ], [ "Export path", 2, "Free egress for leavers on request since 2024-03-05; Velero works; AWS Backup restores only into AWS" ], [ "Stack coupling", 3, "IAM access entries, Pod Identity, VPC CNI, AWS load balancers and EBS storage classes are the default path" ], [ "Hardware / cloud coupling", 2, "Hybrid Nodes and open-source EKS Anywhere and Karpenter run outside AWS, but the managed control plane is AWS only" ], [ "Skills and tooling coupling", 1.5, "kubectl, eksctl and Terraform; AWS IAM knowledge needed" ] ], "integ": [ [ "Partner programme openness", 4, "AWS Partner Network with an EKS service delivery designation; partner count not re-verified" ], [ "Multi-tenancy and self-service", 4, "AWS Organizations and accounts, access entries per cluster, EKS Dashboard across accounts" ], [ "API and infrastructure-as-code", 5, "Official Terraform AWS provider, Cluster API provider, CDK constructs, eksctl and public APIs" ], [ "Skills and certification", 4, "Large practitioner pool and AWS certifications; no EKS-specific certificate" ] ], "alt": null, "ranked": true, "reading": "EKS is building toward hands-off clusters (Auto Mode, Capabilities) and very large AI training estates. The record shows the fastest version cadence of the big three since 2023 and a steady flow of launches, but long-standing roadmap promises can sit for years. A buyer gets depth and scale at a price: no free tier, a sixfold fee in extended support and add-on fees that keep growing.", "report": "reports/01-eks.md", "ai": { "infra": { "2016": null, "2017": null, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 2, "2024": 3, "2025": 3.5, "2026": 3.5 }, "stack": { "2016": null, "2017": null, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 1, "2024": 4, "2025": 4, "2026": 4 }, "ops": { "2016": null, "2017": null, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 1, "2024": 1, "2025": 3, "2026": 3 }, "agents": { "2016": null, "2017": null, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 2, "2024": 2, "2025": 3, "2026": 3 } }, "ai_basis": { "infra": "DRA 2025-07-11, 100,000-node clusters, SOCI lazy image loading 2025-08-29; no managed inference add-on", "stack": "SageMaker HyperPod with EKS orchestration GA 2024-09-10", "ops": "Amazon Q troubleshooting in the EKS console GA 2025-11-21; advisory only", "agents": "Open-source EKS MCP server 2025-05-29; fully managed server in preview from 2025-11-21 with IAM and CloudTrail" }, "ai_ledger": null, "ai_reading": "AI say-do: 1 preview pending (managed EKS MCP server, 10.7 months), 1 roadmap request delivered (DRA); the AI on EKS is HyperPod training orchestration plus advisory troubleshooting, not an acting agent." }, { "key": "aks", "name": "Azure Kubernetes Service (AKS), incl. AKS Automatic and Kubernetes Fleet Manager", "short": "Azure AKS", "kind": "product", "group": "hyperscaler", "since": 2018, "archetype": "Open roadmap, free entry", "grade": "A", "driver": "C", "mix": { "V": 13, "C": 40, "M": 13, "U": 20, "P": 13 }, "idx": 0.2, "domains": { "lifecycle": { "2016": null, "2017": null, "2018": 2, "2019": 2, "2020": 3, "2021": 3, "2022": 3.5, "2023": 4, "2024": 4, "2025": 4.5, "2026": 4.5 }, "compute": { "2016": null, "2017": null, "2018": 2, "2019": 3.5, "2020": 4, "2021": 4, "2022": 4, "2023": 4, "2024": 4, "2025": 4.5, "2026": 4.5 }, "network": { "2016": null, "2017": null, "2018": 2.5, "2019": 3, "2020": 3.5, "2021": 3.5, "2022": 3.5, "2023": 4, "2024": 4.5, "2025": 4.5, "2026": 4.5 }, "storage": { "2016": null, "2017": null, "2018": 2, "2019": 2, "2020": 2, "2021": 3, "2022": 3.5, "2023": 4.5, "2024": 4.5, "2025": 4.5, "2026": 4.5 }, "security": { "2016": null, "2017": null, "2018": 2.5, "2019": 2.5, "2020": 3.5, "2021": 3.5, "2022": 4.5, "2023": 5, "2024": 5, "2025": 5, "2026": 4 }, "ops": { "2016": null, "2017": null, "2018": 2, "2019": 2, "2020": 2, "2021": 2, "2022": 3, "2023": 4, "2024": 4.5, "2025": 4.5, "2026": 4.5 }, "fleet": { "2016": null, "2017": null, "2018": 1, "2019": 1, "2020": 1, "2021": 3, "2022": 3, "2023": 4, "2024": 4.5, "2025": 4.5, "2026": 4.5 }, "devplat": { "2016": null, "2017": null, "2018": 1.5, "2019": 2, "2020": 2, "2021": 2, "2022": 3.5, "2023": 4, "2024": 4, "2025": 4.5, "2026": 4.5 }, "accel": { "2016": null, "2017": null, "2018": 2, "2019": 2, "2020": 2, "2021": 2, "2022": 3.5, "2023": 3.5, "2024": 4, "2025": 4, "2026": 4 }, "sovereignty": { "2016": null, "2017": null, "2018": 2, "2019": 2, "2020": 3, "2021": 3, "2022": 3, "2023": 3.5, "2024": 3.5, "2025": 4, "2026": 4 } }, "cred": [ [ "Cadence", 5, "Weekly dated release notes since 2019, version calendar and release tracker" ], [ "Roadmap transparency", 5, "Public roadmap issues with a stated GA ETA and visible edit history" ], [ "Say-do", 3, "13 of 21 resolved promises GA within 12 months; only 6 of 12 dated ETAs held; Overlay dual-stack for Windows slipped six times" ], [ "Velocity", 5, "Several GA features a month in 2025 and 2026" ], [ "Lifecycle stability", 4, "12 months community support, 24 months with paid LTS, rollback; platform-initiated upgrades after a 60-day grace period" ] ], "comm": [ [ "Price and licence volatility", 2.5, "Standard tier 2023, LTS meter 2023, AKS Automatic $0.16 control-plane fee 2025-10; unexplained $0.05 Free-tier meter from 2026-10-01" ], [ "Purchase constraints", 0.5, "Free control-plane tier, pay as you go" ], [ "Channel and access", 0.5, "Self-service, public docs and pricing, worldwide regions" ], [ "Owner stability", 0.5, "Microsoft, stable and profitable" ], [ "Cost of staying supported", 2.5, "12 months community support; staying longer needs Premium LTS at $0.60 per cluster-hour; platform-initiated upgrades" ] ], "risk": "commercial", "lock": [ [ "Formats", 1.5, "Conformant Kubernetes; Automatic and NAP node classes are provider-specific but workloads use standard APIs" ], [ "Export path", 2, "Egress waived only when leaving Azure completely; Velero works; Azure Backup restores within Azure" ], [ "Stack coupling", 3, "Entra ID, workload identity, Azure CNI, Azure load balancers and Azure storage classes are the default path" ], [ "Hardware / cloud coupling", 1.5, "First-party hybrid through AKS on Azure Local and Arc-enabled Kubernetes" ], [ "Skills and tooling coupling", 1.5, "kubectl, az CLI, Bicep or Terraform" ] ], "integ": [ [ "Partner programme openness", 4, "Microsoft AI Cloud Partner Program with a Kubernetes specialisation; partner count not re-verified" ], [ "Multi-tenancy and self-service", 4, "Management groups and subscriptions, Fleet Manager, managed namespaces" ], [ "API and infrastructure-as-code", 5, "Official azurerm Terraform provider, Bicep and ARM, Cluster API provider for Azure, SDKs" ], [ "Skills and certification", 4, "Large Azure practitioner pool and certifications; no AKS-specific certificate" ] ], "alt": null, "ranked": true, "reading": "AKS is building toward hands-off clusters (Automatic, node auto-provisioning) and fleet-level operations. The record shows a very open roadmap with dated ETAs, but many ETAs slip and big previews take more than a year to reach GA, and 2024 to 2026 brought several provider-side security flaws. A buyer gets a free entry tier, deep Azure integration and LTS, with slower version uptake than EKS.", "report": "reports/02-aks.md", "ai": { "infra": { "2016": null, "2017": null, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 1.5, "2024": 1.5, "2025": 2.5, "2026": 3 }, "stack": { "2016": null, "2017": null, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 2, "2024": 2, "2025": 4, "2026": 4 }, "ops": { "2016": null, "2017": null, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 1.5, "2024": 1.5, "2025": 3.5, "2026": 3.5 }, "agents": { "2016": null, "2017": null, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 2, "2024": 2, "2025": 3, "2026": 3 } }, "ai_basis": { "infra": "Artifact Streaming GA 2026-07-17 for large images; DRA still only proposed", "stack": "KAITO add-on GA (date not confirmed, about 2025) for model inference and fine-tuning", "ops": "Copilot in Azure incl. AKS GA 2025-04-06; agentic CLI in preview since 2025-08", "agents": "Official AKS-MCP server in public preview 2025-08-08, read-only by default" }, "ai_ledger": null, "ai_reading": "AI say-do: 4 items; GPU health GA on its stated date, KAITO GA late, AKS-MCP server and agentic CLI still preview after 14 months; AKS adds KAITO model serving and a portal assistant, the acting agent is not GA." }, { "key": "gke", "name": "Google Kubernetes Engine (Standard and Autopilot; GKE Enterprise edition 2023 to 2025)", "short": "Google GKE", "kind": "product", "group": "hyperscaler", "since": 2016, "archetype": "AI-scale Kubernetes reference", "grade": "A", "driver": "V", "mix": { "V": 27, "C": 27, "M": 13, "U": 13, "P": 20 }, "idx": 0.4, "domains": { "lifecycle": { "2016": 3, "2017": 3, "2018": 3, "2019": 3.5, "2020": 3.5, "2021": 3.5, "2022": 4, "2023": 4, "2024": 4.5, "2025": 4.5, "2026": 4.5 }, "compute": { "2016": 3, "2017": 3, "2018": 3, "2019": 3.5, "2020": 3.5, "2021": 4.5, "2022": 4.5, "2023": 4.5, "2024": 5, "2025": 5, "2026": 5 }, "network": { "2016": 2, "2017": 2, "2018": 3.5, "2019": 3.5, "2020": 3.5, "2021": 4.5, "2022": 4.5, "2023": 4.5, "2024": 5, "2025": 5, "2026": 5 }, "storage": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 2, "2020": 2, "2021": 2, "2022": 3.5, "2023": 4.5, "2024": 4.5, "2025": 5, "2026": 5 }, "security": { "2016": 2, "2017": 2, "2018": 2, "2019": 3, "2020": 3, "2021": 3, "2022": 4.5, "2023": 4.5, "2024": 4.5, "2025": 4, "2026": 4 }, "ops": { "2016": 2, "2017": 2, "2018": 2, "2019": 3, "2020": 3, "2021": 3, "2022": 4, "2023": 4.5, "2024": 4.5, "2025": 4.5, "2026": 4.5 }, "fleet": { "2016": 1, "2017": 1, "2018": 1, "2019": 3.5, "2020": 3.5, "2021": 4.5, "2022": 4.5, "2023": 4.5, "2024": 4.5, "2025": 5, "2026": 5 }, "devplat": { "2016": 2.5, "2017": 2.5, "2018": 2.5, "2019": 3, "2020": 3, "2021": 3, "2022": 4, "2023": 4, "2024": 4, "2025": 4, "2026": 4 }, "accel": { "2016": 0, "2017": 0, "2018": 3, "2019": 3, "2020": 3, "2021": 3, "2022": 4, "2023": 4, "2024": 5, "2025": 5, "2026": 5 }, "sovereignty": { "2016": 3, "2017": 2.5, "2018": 2.5, "2019": 2.5, "2020": 3.5, "2021": 3.5, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 4, "2026": 4 } }, "cred": [ [ "Cadence", 5, "Weekly dated release notes and a published release schedule with channel dates per minor, kept since 2021" ], [ "Roadmap transparency", 3.5, "Release schedule predicts dates by quarter; no public roadmap board, direction comes from Next and KubeCon posts" ], [ "Say-do", 3.5, "10 of 13 resolvable promises 2023-2026 reached GA within 12 months (77%); Gemini Cloud Assist and RayTurbo slipped" ], [ "Velocity", 5, "Several GA capabilities per quarter in 2024-2026: TPUs, DRA, Inference Gateway, Autopilot in Standard, backup features" ], [ "Lifecycle stability", 4, "Extended channel to 24 months since 2024-07; but GKE Enterprise edition launched 2023 and dissolved 2025, No channel removed 2027" ] ], "comm": [ [ "Price and licence volatility", 2.5, "Since 2023: Enterprise edition at $0.00822 per vCPU-hour added then dissolved into SKUs 2025-09; extended fee $0.50 per hour from 2024-07" ], [ "Purchase constraints", 0.5, "Pay as you go, per-second cluster fee, monthly credit covers one zonal or Autopilot cluster" ], [ "Channel and access", 0, "Self-service sign-up, public docs, pricing and release notes, 43 regions" ], [ "Owner stability", 0.5, "Alphabet, profitable and stable; Google Cloud profitable" ], [ "Cost of staying supported", 2, "About 14 months standard support with forced auto-upgrade after; extended support costs $0.50 per cluster-hour" ] ], "risk": "commercial", "lock": [ [ "Formats", 1.5, "CNCF-conformant; Autopilot, ComputeClasses and Config Connector CRDs are Google-specific but optional" ], [ "Export path", 1.5, "Exit data transfer fees waived since 2024-01-11; Backup for GKE is Google-only, Velero works" ], [ "Stack coupling", 2.5, "Workload Identity to Google IAM, Gateway on Cloud Load Balancing, Persistent Disk and Hyperdisk storage classes" ], [ "Hardware / cloud coupling", 1, "Fleet runs GKE on AWS and Azure, attached clusters and Google Distributed Cloud on premises" ], [ "Skills and tooling coupling", 1.5, "kubectl, Terraform and gcloud; Autopilot and fleet concepts are GKE-specific" ] ], "integ": [ [ "Partner programme openness", 4.5, "Google Cloud Partner Advantage with large SI and MSP base; programme details not re-read" ], [ "Multi-tenancy and self-service", 4.5, "Projects, folders and fleet team scopes with namespaces; multi-team management included since 2025-09" ], [ "API and infrastructure-as-code", 5, "hashicorp/google provider 2.65 billion downloads; GKE module 63 million; Config Connector; gcloud and SDKs" ], [ "Skills and certification", 4, "Professional Cloud Architect and DevOps Engineer certifications cover GKE; no GKE-only certification" ] ], "alt": null, "ranked": true, "reading": "Google is building GKE into the place to run very large AI training and inference fleets on its own TPUs and NVIDIA GPUs, while folding Autopilot and fleet features into one product. The release notes show steady weekly delivery, fast Rapid-channel adoption of new Kubernetes minors and most previews reaching GA within a year, but packaging moved twice and the operator AI assistant has stayed in preview since 2023. Buyers get the broadest managed feature set with modest lock-in, at the cost of a per-cluster fee and paid extended support.", "report": "reports/03-gke.md", "ai": { "infra": { "2016": 0, "2017": 0, "2018": 1, "2019": 1, "2020": 1, "2021": 2, "2022": 2, "2023": 3, "2024": 4, "2025": 5, "2026": 5 }, "stack": { "2016": 0, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 2, "2024": 2, "2025": 3, "2026": 3 }, "ops": { "2016": 0, "2017": 0, "2018": 0, "2019": 1, "2020": 1, "2021": 1, "2022": 2, "2023": 2, "2024": 2, "2025": 2.5, "2026": 2.5 }, "agents": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 3, "2026": 4.5 } }, "ai_basis": { "infra": "DRA GA 2025-09, Inference Gateway GA 2025-12, Inference Quickstart, DRANET GA 2026-04", "stack": "Cluster Director for GKE GA at Next 2025, Inference Gateway with model-aware routing", "ops": "Gemini Cloud Assist for GKE in preview since 2023 (half step); investigations read-only, Premium-only from 2026-04", "agents": "Remote GKE MCP server listed without preview label 2026, IAM roles and audit logging documented; tools read-only" }, "ai_ledger": null, "ai_reading": "AI say-do 2023-2026: 8 tracked, 4 GA, 1 partial (Ironwood), 3 not delivered (Duet AI / Gemini Cloud Assist for GKE, Cloud Assist investigations, RayTurbo). The AI that ships is in the workload layer (TPUs, DRA, Inference Gateway); the operator assistant is read-only and still preview." }, { "key": "oke", "name": "Oracle OCI Kubernetes Engine (OKE; Container Engine for Kubernetes until 2024)", "short": "Oracle OKE", "kind": "product", "group": "hyperscaler", "since": 2018, "archetype": "Free-tier GPU cloud Kubernetes", "grade": "A", "driver": "C", "mix": { "V": 0, "C": 45, "M": 27, "U": 9, "P": 18 }, "idx": 0.0, "domains": { "lifecycle": { "2016": null, "2017": null, "2018": 2.5, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 2.5, "2023": 3, "2024": 3, "2025": 3, "2026": 3 }, "compute": { "2016": null, "2017": null, "2018": 2.5, "2019": 2.5, "2020": 2.5, "2021": 3.5, "2022": 3.5, "2023": 4, "2024": 4, "2025": 4, "2026": 4.5 }, "network": { "2016": null, "2017": null, "2018": 2, "2019": 2.5, "2020": 2.5, "2021": 3, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 4, "2026": 4 }, "storage": { "2016": null, "2017": null, "2018": 1.5, "2019": 1.5, "2020": 2, "2021": 2, "2022": 2.5, "2023": 3, "2024": 3, "2025": 4, "2026": 4 }, "security": { "2016": null, "2017": null, "2018": 2, "2019": 2.5, "2020": 2.5, "2021": 3, "2022": 3.5, "2023": 4, "2024": 4, "2025": 4, "2026": 4 }, "ops": { "2016": null, "2017": null, "2018": 1, "2019": 1, "2020": 2, "2021": 2.5, "2022": 2.5, "2023": 3, "2024": 3, "2025": 3, "2026": 3 }, "fleet": { "2016": null, "2017": null, "2018": 1, "2019": 1, "2020": 1, "2021": 1, "2022": 1, "2023": 1, "2024": 1, "2025": 1.5, "2026": 1.5 }, "devplat": { "2016": null, "2017": null, "2018": 2.5, "2019": 2.5, "2020": 2.5, "2021": 3, "2022": 3, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "accel": { "2016": null, "2017": null, "2018": 0, "2019": 0, "2020": 0, "2021": 2.5, "2022": 2.5, "2023": 2.5, "2024": 3, "2025": 3, "2026": 4 }, "sovereignty": { "2016": null, "2017": null, "2018": 2, "2019": 2, "2020": 2.5, "2021": 2.5, "2022": 2.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 } }, "cred": [ [ "Cadence", 4, "Around 25 to 35 dated release notes a year and a versions page with planned and actual dates per patch" ], [ "Roadmap transparency", 2.5, "Planned version dates and preview releases are published; no roadmap board, little direction beyond CloudWorld" ], [ "Say-do", 3, "7 of 10 promises 2023-2026 shipped within 12 months (70%); node health checks took 18 months, two CloudWorld items not found" ], [ "Velocity", 4, "Several features a quarter in 2025-2026: 5,000 then 20,000 nodes, Karpenter, dual stack, rollback, GPU operators" ], [ "Lifecycle stability", 3, "About 12 to 14 months per minor, free in-place upgrades, no extended support; service renamed 2024" ] ], "comm": [ [ "Price and licence volatility", 1, "One change since 2023: paid enhanced tier added 2023-03 while basic clusters stayed free; no later fee changes found" ], [ "Purchase constraints", 0, "Pay as you go; basic clusters free; Free Tier Ampere credits" ], [ "Channel and access", 0.5, "Self-service sign-up, public docs and price API, 50+ regions" ], [ "Owner stability", 2, "Oracle is profitable but carries heavy debt-funded AI data-centre spending (not re-verified from filings in this pass)" ], [ "Cost of staying supported", 2.5, "Minors supported about 12 to 14 months with no extended support option; upgrades free" ] ], "risk": "commercial", "lock": [ [ "Formats", 1, "CNCF-conformant 1.24 to 1.36; few OKE-specific CRDs beyond OCI Service Operator" ], [ "Export path", 1, "Egress 10 TB per month free then $0.0085 per GB; Velero and Cluster API provider available" ], [ "Stack coupling", 2, "Workload identity to OCI IAM, VCN-native CNI and OCI load balancer annotations" ], [ "Hardware / cloud coupling", 3.5, "Runs only on OCI regions; no first-party OKE extension to other clouds or on-premises found" ], [ "Skills and tooling coupling", 1.5, "kubectl and Terraform; OCI compartments and CLI are provider-specific" ] ], "integ": [ [ "Partner programme openness", 3, "Oracle PartnerNetwork with OCI tracks; Kubernetes-specific partner facts not re-read" ], [ "Multi-tenancy and self-service", 3.5, "Compartments and IAM policies per team; no OKE fleet or team-scope feature" ], [ "API and infrastructure-as-code", 4, "oracle/oci Terraform provider 134 million downloads, Oracle-maintained Cluster API provider, SDKs and CLI" ], [ "Skills and certification", 3, "OCI Architect and DevOps certifications; small OKE skills pool" ] ], "alt": null, "ranked": true, "reading": "Oracle is turning OKE into the control layer for very large GPU clusters on OCI, adding node counts, GPU operators and repair tooling at a fast pace since 2025. The record shows steady dated release notes but slow adoption of new Kubernetes minors (about 100 days), no extended support and no fleet or AI-assistant layer. Buyers get a free basic control plane and cheap egress, but a narrower managed feature set than the three largest clouds and coupling to OCI only.", "report": "reports/04-oke.md", "ai": { "infra": { "2016": null, "2017": null, "2018": 0, "2019": 0, "2020": 0, "2021": 1, "2022": 1, "2023": 1, "2024": 2, "2025": 2, "2026": 3 }, "stack": { "2016": null, "2017": null, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 1.5, "2026": 1.5 }, "ops": { "2016": null, "2017": null, "2018": 0.5, "2019": 0.5, "2020": 0.5, "2021": 0.5, "2022": 0.5, "2023": 0.5, "2024": 0.5, "2025": 0.5, "2026": 0.5 }, "agents": { "2016": null, "2017": null, "2018": 1.5, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 2.5, "2026": 2.5 } }, "ai_basis": { "infra": "GPU and Network Operator add-ons, RDMA compute clusters, GPU host repair 2026", "stack": "OCI AI Blueprints installs inference blueprints into OKE; launch date and GA status not verified", "ops": "No AI-assisted operations for OKE found", "agents": "Official open-source oracle/mcp OCI API MCP server 2025-07 runs OCI CLI including OKE; no OKE-specific or managed server" }, "ai_ledger": null, "ai_reading": "AI say-do 2023-2026: 2 tracked, 1 shipped within 12 months (Ubuntu images for GPU workloads), 1 late (node health checks). OKE's AI work is GPU infrastructure plumbing; no operator assistant and no OKE-specific MCP server." }, { "key": "ibm", "name": "IBM Cloud Kubernetes Service and Red Hat OpenShift on IBM Cloud", "short": "IBM IKS / ROKS", "kind": "product", "group": "enterprise", "since": 2017, "archetype": "OpenShift-first hybrid cloud", "grade": "B", "driver": "C", "mix": { "V": 0, "C": 30, "M": 30, "U": 10, "P": 30 }, "idx": 0.0, "domains": { "lifecycle": { "2016": null, "2017": 3, "2018": 3, "2019": 3, "2020": 3, "2021": 3, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "compute": { "2016": null, "2017": 2.5, "2018": 2.5, "2019": 3, "2020": 3, "2021": 3, "2022": 3, "2023": 3, "2024": 3, "2025": 3.5, "2026": 3.5 }, "network": { "2016": null, "2017": 2.5, "2018": 2.5, "2019": 3, "2020": 3, "2021": 3, "2022": 3, "2023": 3, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "storage": { "2016": null, "2017": 2, "2018": 2, "2019": 2, "2020": 2, "2021": 3, "2022": 3, "2023": 3, "2024": 3, "2025": 3.5, "2026": 3.5 }, "security": { "2016": null, "2017": 2.5, "2018": 2.5, "2019": 3, "2020": 3, "2021": 3, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "ops": { "2016": null, "2017": 2, "2018": 2, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3, "2023": 3, "2024": 3, "2025": 3, "2026": 3 }, "fleet": { "2016": null, "2017": 1, "2018": 1, "2019": 1.5, "2020": 1.5, "2021": 3.5, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "devplat": { "2016": null, "2017": 2.5, "2018": 2.5, "2019": 3.5, "2020": 3.5, "2021": 3.5, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "accel": { "2016": null, "2017": 1, "2018": 1, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3, "2023": 3, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "sovereignty": { "2016": null, "2017": 2.5, "2018": 2.5, "2019": 3, "2020": 3, "2021": 3, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 } }, "cred": [ [ "Cadence", 4.5, "Dated release notes several times a week, bi-weekly worker fix packs, version tables with end-of-support dates" ], [ "Roadmap transparency", 2.5, "No roadmap or planned release dates for new minors; direction only through release notes" ], [ "Say-do", 2.5, "4 of 8 dated beta or allowlist items 2023-2026 reached GA within 12 months (50%); H200, Gaudi 3, MI300X, ODF regional DR stalled" ], [ "Velocity", 3, "One or two substantive features a quarter: GPU flavors, secure-by-default networking, OpenShift AI and Virtualization add-ons" ], [ "Lifecycle stability", 3, "IKS minors 14 to 20 months, ROKS about 2 years; Knative add-on and free tier retired" ] ], "comm": [ [ "Price and licence volatility", 1.5, "No control-plane fee changes since 2023, but free cluster tier removed 2023-07-25 and new-account quotas cut from 2024-01-01" ], [ "Purchase constraints", 1, "Pay as you go with no cluster fee; no free tier since 2023; ROKS licence or Cloud Pak entitlement" ], [ "Channel and access", 1, "Self-service sign-up and public docs; prices shown in the console catalogue rather than a public price list" ], [ "Owner stability", 1, "IBM is stable and profitable; IBM Cloud is a smaller cloud with a hybrid and Red Hat focus" ], [ "Cost of staying supported", 1.5, "No extended support, but IKS windows of 14 to 20 months and ROKS about 2 years; upgrades free" ] ], "risk": "commercial", "lock": [ [ "Formats", 1.5, "IKS CNCF-conformant 1.24 to 1.36; ROKS workloads use OpenShift APIs such as Routes and operators" ], [ "Export path", 2, "No migration tooling or exit-egress waiver found; Velero and OADP usable" ], [ "Stack coupling", 2, "IBM Cloud IAM and context-based restrictions, IBM ALBs, Calico settings locked, IBM storage classes" ], [ "Hardware / cloud coupling", 1.5, "Satellite runs ROKS on-premises and on other clouds; ROKS workloads move to other OpenShift platforms" ], [ "Skills and tooling coupling", 2, "ibmcloud CLI and IBM resource groups on top of kubectl or OpenShift skills" ] ], "integ": [ [ "Partner programme openness", 3.5, "IBM Partner Plus and Red Hat partner network; Kubernetes-specific facts not re-read" ], [ "Multi-tenancy and self-service", 3, "Resource groups, IAM access groups and context-based restrictions; 100 clusters per region quota" ], [ "API and infrastructure-as-code", 3.5, "IBM-Cloud/ibm Terraform provider 32.5 million downloads, REST API and CLI plugins" ], [ "Skills and certification", 4, "Red Hat OpenShift certifications widely held; IBM Cloud certifications" ] ], "alt": null, "ranked": true, "reading": "IBM uses ROKS and Satellite to sell managed OpenShift across IBM Cloud, customer sites and other clouds, with IKS as a plain Kubernetes option beside it. The record shows frequent dated release notes and long support windows, but no roadmap, slower OpenShift version adoption and several accelerators stuck on allowlists. Buyers get a free control plane and OpenShift portability, against a smaller managed feature set and a narrower cloud footprint than the hyperscalers.", "report": "reports/05-ibm.md", "ai": { "infra": { "2016": null, "2017": 0, "2018": 0, "2019": 1, "2020": 1, "2021": 1, "2022": 1, "2023": 1, "2024": 2, "2025": 2, "2026": 2 }, "stack": { "2016": null, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 3, "2026": 3 }, "ops": { "2016": null, "2017": 0.5, "2018": 0.5, "2019": 0.5, "2020": 0.5, "2021": 0.5, "2022": 0.5, "2023": 0.5, "2024": 0.5, "2025": 0.5, "2026": 0.5 }, "agents": { "2016": null, "2017": 1.5, "2018": 1.5, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 2, "2026": 2 } }, "ai_basis": { "infra": "L4, L40 and H100 flavors GA 2024; newer accelerators allowlist only; no GPU sharing or inference add-on", "stack": "Red Hat OpenShift AI add-on for ROKS 2025-01-21, deployable at cluster creation 2025-02", "ops": "No AI-assisted operations for IKS or ROKS verified", "agents": "IBM MCP catalogue 2025 lists a Terraform IBM Modules MCP server; no IKS or ROKS MCP server" }, "ai_ledger": null, "ai_reading": "AI say-do 2023-2026: 4 accelerator items tracked, 1 GA within 12 months (H100), 3 still allowlist (H200, Gaudi 3, MI300X). The AI story runs through OpenShift AI on ROKS; there is no operator assistant or MCP server for the Kubernetes services." }, { "key": "rosa", "name": "Red Hat OpenShift Service on AWS (ROSA), Azure Red Hat OpenShift (ARO) and OpenShift Dedicated", "short": "Red Hat managed OpenShift", "kind": "product", "group": "enterprise", "since": 2016, "archetype": "Enterprise OpenShift as a cloud service", "grade": "C", "driver": "C", "mix": { "V": 8, "C": 58, "M": 17, "U": 8, "P": 8 }, "idx": 0.1, "domains": { "lifecycle": { "2016": 2, "2017": 2, "2018": 2, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3, "2023": 3, "2024": 3, "2025": 3.5, "2026": 3.5 }, "compute": { "2016": 2, "2017": 2, "2018": 2, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3, "2023": 3, "2024": 3.5, "2025": 3.5, "2026": 4 }, "network": { "2016": 2, "2017": 2, "2018": 2, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3, "2023": 3, "2024": 3, "2025": 3.5, "2026": 3.5 }, "storage": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 2, "2020": 2, "2021": 2, "2022": 2.5, "2023": 2.5, "2024": 3, "2025": 3, "2026": 3 }, "security": { "2016": 2, "2017": 2, "2018": 2, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3, "2023": 3, "2024": 4, "2025": 4, "2026": 3.5 }, "ops": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3, "2023": 3, "2024": 3, "2025": 3, "2026": 3.5 }, "fleet": { "2016": 1, "2017": 1, "2018": 1, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 2.5 }, "devplat": { "2016": 3, "2017": 3, "2018": 3, "2019": 3.5, "2020": 3.5, "2021": 4, "2022": 4, "2023": 4, "2024": 4, "2025": 4, "2026": 4 }, "accel": { "2016": 0, "2017": 0, "2018": 0, "2019": 0.5, "2020": 0.5, "2021": 0.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 3, "2026": 3.5 }, "sovereignty": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 2.5, "2020": 3, "2021": 3, "2022": 3, "2023": 3, "2024": 3.5, "2025": 3.5, "2026": 3.5 } }, "cred": [ [ "Cadence", 4, "OCP minor every four months; ROSA quarterly and ARO monthly dated release notes, published life cycle tables" ], [ "Roadmap transparency", 3, "Dated release notes and press releases; no public roadmap board for ROSA or ARO" ], [ "Say-do", 3, "4 of 6 resolved promises shipped within about 12 months; Lightspeed promised for late 2024 shipped 2025-05-20" ], [ "Velocity", 3.5, "Several ROSA features most quarters in 2025-26: egress zero, capacity blocks, Karpenter, Spot; ARO slower" ], [ "Lifecycle stability", 3.5, "18 months per minor plus included EUS on even minors; forced upgrades at end of life; ARO installs four to eight months late" ] ], "comm": [ [ "Price and licence volatility", 1.5, "ROSA classic $0.03 cluster fee dropped and ROSA HCP $0.25 per hour cluster fee added around 2024-01; worker fee unchanged since 2021" ], [ "Purchase constraints", 2, "Pay as you go through AWS or Azure, but no free tier and a per-vCPU service fee on every worker" ], [ "Channel and access", 0.5, "Self-service in AWS and Azure consoles, public docs and pricing, about 32 AWS regions" ], [ "Owner stability", 1, "Red Hat under IBM since 2019, stable; service depends on joint operation with AWS and Microsoft" ], [ "Cost of staying supported", 1.5, "18 months per minor with EUS included, in-place upgrades free; forced upgrades at end of life" ] ], "risk": "commercial", "lock": [ [ "Formats", 2.5, "Certified Kubernetes, but many workloads use OpenShift Routes, ImageStreams and security context constraints" ], [ "Export path", 2, "OADP (Velero) documented; egress at AWS or Azure rates; no free leaver egress found" ], [ "Stack coupling", 2, "AWS STS roles or Entra identities, cloud load balancers and storage classes, with OpenShift layers above them" ], [ "Hardware / cloud coupling", 0.5, "Same OpenShift runs self-managed on-premises, on IBM Cloud and on other clouds" ], [ "Skills and tooling coupling", 2.5, "OpenShift-specific oc, operator and rosa CLI skills; widely taught but not plain kubectl" ] ], "integ": [ [ "Partner programme openness", 4.5, "Red Hat partner ecosystem plus AWS and Microsoft channels; launch partners Accenture, Cognizant, Persistent" ], [ "Multi-tenancy and self-service", 3.5, "Hybrid Cloud Console organisations, billing accounts and projects; multicluster only through separately licensed ACM" ], [ "API and infrastructure-as-code", 4, "Partner Terraform provider rhcs with about 1.0M downloads, ARO in the Azure provider, rosa CLI and OCM API" ], [ "Skills and certification", 4.5, "Red Hat OpenShift certifications are widely held; certification details not re-checked this run" ] ], "alt": null, "ranked": true, "reading": "Red Hat sells the same OpenShift platform as a managed service on AWS, Azure and Google Cloud, moving ROSA to hosted control planes and adding Karpenter, Spot and GPU capacity options. The record shows steady quarterly ROSA delivery with versions about six months behind upstream, while ARO trails by about a year and only reached hosted control planes in preview in 2026. Buyers gain one platform across clouds and on-premises with long included support, at the price of per-vCPU fees and OpenShift-specific skills and APIs.", "report": "reports/06-rosa.md", "ai": { "infra": { "2016": 0, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 1, "2024": 1, "2025": 2, "2026": 2 }, "stack": { "2016": 0, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 1, "2022": 1, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 2.5 }, "ops": { "2016": 0, "2017": 0, "2018": 0, "2019": 1, "2020": 1, "2021": 1, "2022": 1, "2023": 1, "2024": 1, "2025": 3, "2026": 3 }, "agents": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 1.5, "2020": 1.5, "2021": 2, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 3 } }, "ai_basis": { "infra": "Capacity Blocks for ML and Trainium and Inferentia instance types on ROSA", "stack": "Red Hat OpenShift AI announced 2023-05-23, runs on ROSA and ARO as a separate subscription", "ops": "OpenShift Lightspeed GA 2025-05-20, natural-language console assistant; cluster interaction still preview", "agents": "MCP server for OpenShift in technology preview 2026-05-05, read-only default with OIDC and RBAC" }, "ai_ledger": null, "ai_reading": "AI say-do: 3 announced, 1 GA (Lightspeed, about six months late), 2 preview pending (cluster interaction, MCP server), 0 dropped; AI work is a bring-your-own-model console assistant plus OpenShift AI sold separately." }, { "key": "ack", "name": "Alibaba Cloud Container Service for Kubernetes (ACK Pro and Basic, ACK Serverless / ACS, ACK One)", "short": "Alibaba ACK", "kind": "product", "group": "asia", "since": 2017, "archetype": "Fast-moving Asian AI-infrastructure cloud", "grade": "A", "driver": "C", "mix": { "V": 18, "C": 27, "M": 27, "U": 18, "P": 9 }, "idx": 0.25, "domains": { "lifecycle": { "2016": null, "2017": 1.5, "2018": 1.5, "2019": 3, "2020": 3, "2021": 3, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 4, "2026": 4 }, "compute": { "2016": null, "2017": 2, "2018": 2, "2019": 3.5, "2020": 3.5, "2021": 3.5, "2022": 4, "2023": 4, "2024": 4, "2025": 4.5, "2026": 4.5 }, "network": { "2016": null, "2017": 2, "2018": 2, "2019": 3, "2020": 3, "2021": 3, "2022": 4, "2023": 4, "2024": 4, "2025": 4.5, "2026": 4.5 }, "storage": { "2016": null, "2017": 1, "2018": 1, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 4 }, "security": { "2016": null, "2017": 1.5, "2018": 1.5, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 4, "2026": 4 }, "ops": { "2016": null, "2017": 1.5, "2018": 1.5, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3.5, "2023": 3.5, "2024": 4, "2025": 4, "2026": 4 }, "fleet": { "2016": null, "2017": 0.5, "2018": 0.5, "2019": 1, "2020": 1, "2021": 3.5, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 4.5, "2026": 4.5 }, "devplat": { "2016": null, "2017": 1.5, "2018": 1.5, "2019": 3, "2020": 3, "2021": 3, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 4 }, "accel": { "2016": null, "2017": 0.5, "2018": 0.5, "2019": 2, "2020": 2, "2021": 2, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 4.5, "2026": 4.5 }, "sovereignty": { "2016": null, "2017": 1, "2018": 1, "2019": 2, "2020": 3, "2021": 3, "2022": 3, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 } }, "cred": [ [ "Cadence", 4.5, "Monthly English release notes since 2017 and a published version table with end-of-support dates" ], [ "Roadmap transparency", 3, "Dated release notes but no public roadmap; some detail only on Chinese pages" ], [ "Say-do", 3, "Few dated promises: version-timing policy kept for 1.31 to 1.36; MCP server near-term tools still missing after 12 months" ], [ "Velocity", 4.5, "Several substantive features most months: Auto Mode, DRA, inference gateway, hybrid node pools, Argo cluster" ], [ "Lifecycle stability", 3, "12-month windows from 1.31, forced updates, no extended support; dedicated clusters retired and serverless renamed twice" ] ], "comm": [ [ "Price and licence volatility", 1.5, "Pro fee since 2020, Basic limited to 2 clusters and 10 nodes, ASK renamed ACK Serverless 2023 then ACS; AI suite made free 2024" ], [ "Purchase constraints", 1, "Pay as you go, free Basic tier for testing; identity verification needed" ], [ "Channel and access", 1.5, "Self-service international site with English docs and SLA; some details Chinese only" ], [ "Owner stability", 1.5, "Alibaba Group is profitable and listed; cross-border regulatory exposure for non-Chinese buyers" ], [ "Cost of staying supported", 2.5, "12 months per minor from 1.31, forced updates one month after notice, no paid extended support" ] ], "risk": "commercial", "lock": [ [ "Formats", 1.5, "CNCF-certified for every offered minor; optional ACK CRDs for quotas, Fluid and Auto Mode" ], [ "Export path", 1.5, "Backup center restores applications and volumes to self-managed clusters; egress at Alibaba rates" ], [ "Stack coupling", 2.5, "Terway ENI networking, RAM and RRSA identity, SLB and ALB, CNFS storage abstractions" ], [ "Hardware / cloud coupling", 1.5, "ACK One registers external and on-premises clusters; hybrid node pools add data-center nodes" ], [ "Skills and tooling coupling", 1.5, "Standard kubectl and Terraform; aliyun CLI and console for cloud parts" ] ], "integ": [ [ "Partner programme openness", 3, "Alibaba Cloud partner programme exists; not re-checked this run" ], [ "Multi-tenancy and self-service", 3.5, "Up to 100 Pro clusters per account, RAM and RBAC, ACK One fleets across clusters" ], [ "API and infrastructure-as-code", 4.5, "Partner Terraform provider aliyun/alicloud with 68.8M downloads, OpenAPI, CLI and SDKs, ACK One Go SDK" ], [ "Skills and certification", 3, "Alibaba ACP and ACE certifications exist; less common outside Asia; not re-checked" ] ], "alt": null, "ranked": true, "reading": "Alibaba is turning ACK into an automated, AI-centred platform: Auto Mode nodes, Lingjun AI clusters, inference routing on Gateway API and an LLM assistant with an MCP server. The release notes show a feature or more almost every month and new Kubernetes minors within weeks, but there is no public roadmap and support windows were cut to 12 months with forced updates. Buyers get very broad capability and fleet tooling at low control-plane cost, weighed against regional reach outside Asia, sparse compliance evidence and a provider whose roadmap is opaque.", "report": "reports/07-ack.md", "ai": { "infra": { "2016": null, "2017": 0, "2018": 0, "2019": 1, "2020": 1, "2021": 1, "2022": 2.5, "2023": 3.5, "2024": 3.5, "2025": 4.5, "2026": 4.5 }, "stack": { "2016": null, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 1.5, "2023": 1.5, "2024": 2.5, "2025": 3, "2026": 3 }, "ops": { "2016": null, "2017": 0, "2018": 0, "2019": 1, "2020": 1, "2021": 1, "2022": 1, "2023": 1, "2024": 3, "2025": 3, "2026": 3.5 }, "agents": { "2016": null, "2017": 1.5, "2018": 1.5, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 4, "2026": 4 } }, "ai_basis": { "infra": "Gateway with Inference Extension 2025-03 and KV-cache-aware routing 2025-08", "stack": "Kagent, AgentScope on Knative and routing to Bailian model services 2025-10 to 2026-03", "ops": "AI Assistant agent mode for root-cause analysis 2026-02", "agents": "Official ack-mcp-server v1.0.0 2025-11-13 with read-write control; enterprise RBAC still on roadmap" }, "ai_ledger": null, "ai_reading": "AI say-do: 2 announced (MCP lifecycle tools, MCP RBAC), 0 GA, 2 pending, 0 dropped; the AI assistant, its agent mode and the MCP server shipped without prior promises, and inference routing on Gateway API is the strongest AI workload feature." }, { "key": "tke", "name": "Tencent Kubernetes Engine (TKE general clusters with native and super nodes)", "short": "Tencent TKE", "kind": "product", "group": "asia", "since": 2016, "archetype": "Conservative Asian cloud Kubernetes", "grade": "A", "driver": "C", "mix": { "V": 0, "C": 50, "M": 20, "U": 10, "P": 20 }, "idx": 0.0, "domains": { "lifecycle": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 3, "2020": 3, "2021": 3, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "compute": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 3, "2020": 3, "2021": 3, "2022": 4, "2023": 4, "2024": 4, "2025": 4, "2026": 4 }, "network": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 2.5, "2020": 3.5, "2021": 3.5, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "storage": { "2016": 1, "2017": 1, "2018": 1, "2019": 2.5, "2020": 3, "2021": 3, "2022": 3, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "security": { "2016": 1, "2017": 1, "2018": 1, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3, "2023": 3, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "ops": { "2016": 1, "2017": 1, "2018": 1, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 4, "2026": 4 }, "fleet": { "2016": 0.5, "2017": 0.5, "2018": 0.5, "2019": 1, "2020": 1, "2021": 2.5, "2022": 2.5, "2023": 3, "2024": 3, "2025": 3, "2026": 3 }, "devplat": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 3, "2020": 3, "2021": 3.5, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "accel": { "2016": 0, "2017": 0, "2018": 0, "2019": 2, "2020": 2, "2021": 3.5, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "sovereignty": { "2016": 1, "2017": 1, "2018": 1, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3, "2023": 3, "2024": 3, "2025": 3, "2026": 3 } }, "cred": [ [ "Cadence", 3.5, "Dated monthly English update notes 2016 to mid-2025 and a release schedule table; English notes stop after June 2025" ], [ "Roadmap transparency", 2.5, "No public roadmap; English documentation lags and is partly machine-translated" ], [ "Say-do", 2, "Version policy promises GA about three months after upstream; met for 1 of 5 minors (1.28 at 257 days)" ], [ "Velocity", 3, "Steady incremental node, storage and policy features; few large launches since 2023" ], [ "Lifecycle stability", 2.5, "27 months per even minor, but serverless clusters, qGPU on normal nodes, TencentHub, TPS and NginxIngress updates withdrawn" ] ], "comm": [ [ "Price and licence volatility", 3, "Free managed clusters became paid by cluster model on 2022-03-21, quotas imposed 2022-04-30, serverless clusters withdrawn 2025" ], [ "Purchase constraints", 1.5, "Pay as you go from USD 0.02 per hour; no free managed tier" ], [ "Channel and access", 1.5, "Self-service international site with English docs, some machine-translated; script access blocked on the main doc domain" ], [ "Owner stability", 1.5, "Tencent is profitable and listed; cross-border regulatory exposure for non-Chinese buyers" ], [ "Cost of staying supported", 1.5, "27-month support included; forced upgrades possible after expiry; even minors only" ] ], "risk": "commercial", "lock": [ [ "Formats", 1.5, "CNCF-certified for each released even minor; workloads use standard APIs" ], [ "Export path", 2, "Backup Center with cross-cluster restore since 2024-04; egress at Tencent rates" ], [ "Stack coupling", 2.5, "VPC-CNI, CAM, CLB and CBS integration; TKE re-creates cloud resources declaratively" ], [ "Hardware / cloud coupling", 2, "Registered clusters and registered IDC nodes; no first-party on-premises distribution counted here" ], [ "Skills and tooling coupling", 1.5, "Standard kubectl and Terraform; tccli for cloud parts" ] ], "integ": [ [ "Partner programme openness", 2.5, "Partner programme not checked this run; limited presence outside Asia" ], [ "Multi-tenancy and self-service", 3, "CAM, projects and tags; registered clusters; no fleet layer" ], [ "API and infrastructure-as-code", 4.5, "Partner Terraform provider tencentcloudstack/tencentcloud with 46M downloads, Cloud API 3.0, tccli" ], [ "Skills and certification", 2.5, "Tencent Cloud certifications exist; not checked this run" ] ], "alt": null, "ranked": true, "reading": "Tencent runs TKE as a stable, even-minor Kubernetes service with long 27-month support and incremental node, storage and policy features, consolidating serverless into super nodes. The record shows versions arriving well after its own three-month target and no public roadmap or AI operations tooling, while managed clusters became paid in 2022. Buyers in Asia get a mature, inexpensive service with long support, weighed against slow version adoption, thin English documentation and limited reach elsewhere.", "report": "reports/08-tke.md", "ai": { "infra": { "2016": 0, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 2, "2022": 2, "2023": 2, "2024": 2, "2025": 2.5, "2026": 2.5 }, "stack": { "2016": 0, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 1, "2022": 1, "2023": 1, "2024": 1, "2025": 1, "2026": 1 }, "ops": { "2016": 0, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 1.5 }, "agents": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 2, "2020": 2, "2021": 2, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 2 } }, "ai_basis": { "infra": "AIBrix distributed inference and LLM deployment guides 2025-04, open-source components", "stack": "Cloud-native AI service 2021-10; TI platform integration not confirmed", "ops": "Rule-based health checks and Request Recommendation 2021-10; no AI assistant found", "agents": "Cloud API 3.0 with all-region access and tccli; no official MCP server found" }, "ai_ledger": null, "ai_reading": "AI say-do: 0 announced, 0 GA, 0 preview, 0 pending, 0 dropped; TKE's AI offer is GPU sharing plus documented open-source inference stacks, with no operator assistant or MCP server found in English docs." }, { "key": "cce", "name": "Huawei Cloud Cloud Container Engine (CCE standard and Turbo, CCE Autopilot, with UCS fleet management)", "short": "Huawei CCE", "kind": "product", "group": "asia", "since": 2016, "archetype": "Hybrid-minded Asian cloud with own accelerators", "grade": "C", "driver": "C", "mix": { "V": 20, "C": 30, "M": 20, "U": 20, "P": 10 }, "idx": 0.29, "domains": { "lifecycle": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 3, "2020": 3, "2021": 3, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "compute": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 3, "2020": 3, "2021": 3, "2022": 3.5, "2023": 3.5, "2024": 4, "2025": 4, "2026": 4 }, "network": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 4, "2026": 4 }, "storage": { "2016": 1, "2017": 1, "2018": 1, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3, "2023": 3, "2024": 3, "2025": 3, "2026": 4 }, "security": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 2, "2020": 2, "2021": 2, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 4, "2026": 4 }, "ops": { "2016": 1, "2017": 1, "2018": 1, "2019": 2, "2020": 2, "2021": 2, "2022": 3, "2023": 3, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "fleet": { "2016": 0.5, "2017": 0.5, "2018": 0.5, "2019": 1, "2020": 1, "2021": 1, "2022": 1.5, "2023": 4, "2024": 4, "2025": 4.5, "2026": 4.5 }, "devplat": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3, "2023": 3, "2024": 3, "2025": 3.5, "2026": 3.5 }, "accel": { "2016": 0.5, "2017": 0.5, "2018": 0.5, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 4, "2026": 4 }, "sovereignty": { "2016": 1, "2017": 1, "2018": 1, "2019": 2, "2020": 2, "2021": 2, "2022": 3, "2023": 3, "2024": 3, "2025": 3.5, "2026": 3.5 } }, "cred": [ [ "Cadence", 4, "Monthly What's New with OBT or commercial phase per item, version policy table, dated notices" ], [ "Roadmap transparency", 3, "Public OBT phases signal upcoming GA, but no roadmap with dates" ], [ "Say-do", 4, "12 of 13 resolved OBT items reached commercial use within 1 to 5 months; DataPlane V2 on Turbo still in OBT after 19 months" ], [ "Velocity", 4, "Several features a month in 2025-26: DataPlane V2, Cluster Backup, Gateway API, AI gateway, Autopilot" ], [ "Lifecycle stability", 3.5, "24 months per minor with EOM notices ahead; Docker, EulerOS 2.9 and NGINX Ingress policies changed 2025" ] ], "comm": [ [ "Price and licence volatility", 1, "No fee or tier change found since 2023; cluster fee by scale and HA" ], [ "Purchase constraints", 1.5, "Pay-per-use per second or yearly/monthly; no free cluster tier found" ], [ "Channel and access", 1.5, "Self-service international site with English docs and SLA" ], [ "Owner stability", 2.5, "Privately held Huawei under US export restrictions; procurement limits in some markets" ], [ "Cost of staying supported", 1.5, "24 months per minor included, upgrades free; no SLA after EOS" ] ], "risk": "commercial", "lock": [ [ "Formats", 1.5, "CNCF-certified for each released minor; standard APIs" ], [ "Export path", 1.5, "Velero-based Cluster Backup supports migration since 2026-06; egress at Huawei rates" ], [ "Stack coupling", 2.5, "CCE Turbo ENI passthrough, ELB annotations, everest storage classes and IAM" ], [ "Hardware / cloud coupling", 1, "UCS runs on-premises clusters and attaches clusters elsewhere" ], [ "Skills and tooling coupling", 1.5, "Standard kubectl and Terraform; hcloud CLI for cloud parts" ] ], "integ": [ [ "Partner programme openness", 2.5, "Partner programme not checked; reach outside China limited by restrictions" ], [ "Multi-tenancy and self-service", 4, "IAM enterprise projects for nodes and PVCs, UCS fleets and federation RBAC" ], [ "API and infrastructure-as-code", 3.5, "Terraform provider huaweicloud/huaweicloud with 4.2M downloads (community tier), APIs and CLI" ], [ "Skills and certification", 3, "Huawei HCIA and HCIP cloud certifications exist; not checked this run" ] ], "alt": null, "ranked": true, "reading": "Huawei is building CCE into a hybrid platform: UCS fleets reach on-premises and other clusters, Autopilot adds serverless clusters and the AI Suite schedules both NVIDIA GPUs and its own Ascend NPUs. Its release notes label every preview and show most previews reaching commercial use within months, while Kubernetes versions arrive about six months after upstream with 24 months of support. Buyers gain strong fleet and accelerator options and clear phases, weighed against slower versions, sparse pricing and compliance evidence in English, and procurement limits tied to Huawei's export restrictions.", "report": "reports/09-cce.md", "ai": { "infra": { "2016": 0, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 2, "2024": 2, "2025": 3, "2026": 3.5 }, "stack": { "2016": 0, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 1, "2024": 1, "2025": 1, "2026": 1.5 }, "ops": { "2016": 0, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 1, "2023": 1, "2024": 1, "2025": 1, "2026": 1 }, "agents": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 3 } }, "ai_basis": { "infra": "Envoy AI Inference Gateway add-on commercial 2026-06", "stack": "AgentCube and Notebook add-ons in OBT 2026", "ops": "Rule-based npd fault detection; no AI assistant found", "agents": "Official Huawei Cloud devkit MCP covering CCE, beta with write guardrails, 2026-07" }, "ai_ledger": null, "ai_reading": "AI say-do: 3 announced (AgentCube, Notebook, devkit MCP), 0 GA, 3 preview pending, 0 dropped; AI strength is Ascend and GPU scheduling plus an inference gateway, with no operator assistant found in English docs." }, { "key": "doks", "name": "DigitalOcean Kubernetes (DOKS)", "short": "DigitalOcean DOKS", "kind": "product", "group": "developer", "since": 2019, "archetype": "Simple developer cloud", "grade": "B", "driver": "C", "mix": { "V": 10, "C": 60, "M": 20, "U": 10, "P": 0 }, "idx": 0.11, "domains": { "lifecycle": { "2016": null, "2017": null, "2018": null, "2019": 2.5, "2020": 2.5, "2021": 3, "2022": 3.5, "2023": 3.5, "2024": 4, "2025": 4, "2026": 4 }, "compute": { "2016": null, "2017": null, "2018": null, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 3, "2026": 3 }, "network": { "2016": null, "2017": null, "2018": null, "2019": 2.5, "2020": 3, "2021": 3, "2022": 3, "2023": 3, "2024": 3, "2025": 3.5, "2026": 3.5 }, "storage": { "2016": null, "2017": null, "2018": null, "2019": 2, "2020": 2, "2021": 2, "2022": 2, "2023": 2, "2024": 2.5, "2025": 3, "2026": 3 }, "security": { "2016": null, "2017": null, "2018": null, "2019": 2, "2020": 2, "2021": 2, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 3 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 2.5, "2023": 2.5, "2024": 3, "2025": 3, "2026": 3 }, "fleet": { "2016": null, "2017": null, "2018": null, "2019": 2, "2020": 2, "2021": 2, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 2 }, "devplat": { "2016": null, "2017": null, "2018": null, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 2.5 }, "accel": { "2016": null, "2017": null, "2018": null, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 3, "2025": 3.5, "2026": 3.5 }, "sovereignty": { "2016": null, "2017": null, "2018": null, "2019": 2, "2020": 2, "2021": 2, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 3 } }, "cred": [ [ "Cadence", 4, "Monthly patch releases and dated version table; 1.37 released 2026-10-08, 43 days after upstream" ], [ "Roadmap transparency", 2.5, "Dated release notes and changelog; no public roadmap board found" ], [ "Say-do", 3.5, "11 of 13 judged previews reached GA within 12 months; control plane firewall took 642 days; one beta without GA" ], [ "Velocity", 4, "Many GA features in 2025 and 2026: Gateway API, NFS, SSO, isolated nodes, GPU nodes" ], [ "Lifecycle stability", 3, "About 14 months per minor, no extended support, forced upgrade 30 days after end of support" ] ], "comm": [ [ "Price and licence volatility", 1.5, "No fee change since the 2022 HA price; HA on by default from 1.36 with unclear billing; IPv4 line item for v5 nodes" ], [ "Purchase constraints", 1, "Self-service pay as you go; newest GPUs by contract only" ], [ "Channel and access", 0.5, "Self-service, public docs and pricing, 16 datacenter codes" ], [ "Owner stability", 1.5, "Listed company; Paperspace bought 2023 for $111 million; share offering in 2026-03" ], [ "Cost of staying supported", 2, "Upgrades free but forced 30 days after end of support; no extended support" ] ], "risk": "commercial", "lock": [ [ "Formats", 0.5, "Certified Kubernetes (folders 1.24 to 1.36); standard APIs with load balancer annotations" ], [ "Export path", 1.5, "No migration service found; outbound transfer $0.01 per GiB above allowance" ], [ "Stack coupling", 1.5, "DigitalOcean load balancers, volumes and firewalls reconciled by the service" ], [ "Hardware / cloud coupling", 3.5, "DigitalOcean cloud only; no hybrid or on-premises extension found" ], [ "Skills and tooling coupling", 0.5, "kubectl, Terraform and doctl" ] ], "integ": [ [ "Partner programme openness", 2, "Terraform partner-tier provider; partner programme not verified" ], [ "Multi-tenancy and self-service", 2, "Teams and Projects; no organisation or fleet isolation found" ], [ "API and infrastructure-as-code", 3.5, "Terraform provider 15.7 million downloads, doctl, API and a community Cluster API provider" ], [ "Skills and certification", 2, "No Kubernetes certification path found" ] ], "alt": null, "ranked": true, "reading": "DOKS is building toward GPU and private-network features on a simple, cheap base. The record shows fast preview-to-GA moves since 2024 but no roadmap, no fleet tooling and no extended support. A buyer gets low cost and standard Kubernetes with a single cloud and limited enterprise depth.", "report": "reports/10-doks.md", "ai": { "infra": { "2016": null, "2017": null, "2018": null, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 2, "2025": 2.5, "2026": 2.5 }, "stack": { "2016": null, "2017": null, "2018": null, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 0.5, "2024": 0.5, "2025": 0.5, "2026": 0.5 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": 1, "2020": 1, "2021": 1, "2022": 1, "2023": 1, "2024": 1, "2025": 1, "2026": 1 }, "agents": { "2016": null, "2017": null, "2018": null, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 3, "2026": 3 } }, "ai_basis": { "infra": "Multi-node 8-GPU clusters, GPU metrics 2025-11-03; DRA preview 2026-08-13", "stack": "Gradient and Paperspace are separate products; no DOKS integration found", "ops": "Rule-based insights and alerts", "agents": "Remote MCP servers 2025-12-05 with a DOKS tool group and OAuth; local server is a Labs repository" }, "ai_ledger": null, "ai_reading": "AI say-do: 1 early access reached GA (GPU nodes, 32 days), 3 previews pending; DOKS offers GPU pools and MCP tools but no assistant or integrated AI platform." }, { "key": "lke", "name": "Akamai Linode Kubernetes Engine (LKE, LKE Enterprise)", "short": "Akamai LKE", "kind": "product", "group": "developer", "since": 2020, "archetype": "Akamai developer cloud", "grade": "B", "driver": "C", "mix": { "V": 11, "C": 33, "M": 22, "U": 22, "P": 11 }, "idx": 0.17, "domains": { "lifecycle": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 2, "2021": 2, "2022": 2, "2023": 2, "2024": 2.5, "2025": 2.5, "2026": 2.5 }, "compute": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 2, "2021": 2, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 3, "2026": 3 }, "network": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 2, "2021": 2, "2022": 2, "2023": 2, "2024": 2, "2025": 2.5, "2026": 2.5 }, "storage": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 2, "2021": 2, "2022": 2, "2023": 2, "2024": 2, "2025": 2.5, "2026": 2.5 }, "security": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 2.5 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 2, "2025": 2, "2026": 2 }, "fleet": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1, "2021": 1, "2022": 1, "2023": 1, "2024": 1, "2025": 1.5, "2026": 1.5 }, "devplat": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 2, "2026": 2 }, "accel": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 2.5, "2026": 2.5 }, "sovereignty": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1.5, "2021": 1.5, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 2.5 } }, "cred": [ [ "Cadence", 3.5, "Release notes every few weeks since 2023-07; patch versions bumped regularly" ], [ "Roadmap transparency", 2.5, "Dated release notes and lifecycle table; no public roadmap board" ], [ "Say-do", 2, "3 of 7 signals reached GA within 12 months; LKE Enterprise still limited; Enterprise minors 314 to 485 days late" ], [ "Velocity", 3, "Steady GPU, logging and IAM additions in 2025 and 2026, fewer Kubernetes-specific features" ], [ "Lifecycle stability", 2.5, "About 12 months per minor on LKE; forced upgrades; Enterprise lags upstream by three minors" ] ], "comm": [ [ "Price and licence volatility", 1.5, "HA price stable at $60 since 2022; new $300 Enterprise tier; hourly billing change for GPU and G8 in 2026-07" ], [ "Purchase constraints", 1.5, "Standard LKE self-service; LKE Enterprise and capacity need sales" ], [ "Channel and access", 1.5, "Docs and pricing public; 33 regions; Enterprise limited availability" ], [ "Owner stability", 1.5, "Large listed owner, but a 2022 acquisition and a shift to AI cloud strategy" ], [ "Cost of staying supported", 2.5, "Three minors only, EOL about one year, forced upgrades on standard" ] ], "risk": "commercial", "lock": [ [ "Formats", 0.5, "Certified Kubernetes (folders 1.24 to 1.36, none for 1.29); standard APIs" ], [ "Export path", 1.5, "Overage egress $0.005 per GB; no migration service found" ], [ "Stack coupling", 1.5, "NodeBalancers, Block Storage CSI and Cloud Firewall coupling" ], [ "Hardware / cloud coupling", 3.5, "Akamai cloud only; no hybrid extension found" ], [ "Skills and tooling coupling", 0.5, "kubectl and Terraform" ] ], "integ": [ [ "Partner programme openness", 2, "Terraform partner-tier provider; partner programme not verified" ], [ "Multi-tenancy and self-service", 2, "Account users and IAM RBAC GA 2026-03-31; no fleet or org isolation" ], [ "API and infrastructure-as-code", 4, "Terraform provider 18.6 million downloads, API, CLI and a Cluster API provider" ], [ "Skills and certification", 2, "No Kubernetes certification path found" ] ], "alt": null, "ranked": true, "reading": "LKE is moving from a simple developer service toward an enterprise tier and GPU cloud under Akamai. The record shows steady maintenance and new GPU plans, but LKE Enterprise is still limited, runs three minors behind upstream and has no roadmap board. A buyer gets a cheap, standard cluster with few enterprise add-ons.", "report": "reports/11-lke.md", "ai": { "infra": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 2, "2026": 2.5 }, "stack": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 1, "2025": 1, "2026": 1 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 0.5, "2021": 0.5, "2022": 0.5, "2023": 0.5, "2024": 0.5, "2025": 0.5, "2026": 0.5 }, "agents": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 2.5 } }, "ai_basis": { "infra": "Blackwell 8-card plans 2026-07-31; DRA on Enterprise 1.34", "stack": "App Platform beta 2024-11-13 with KServe and Kubeflow; no GA date", "ops": "Docs only", "agents": "Official read-only MCP server, version 0.1.0, 2026-06-13; cannot change clusters" }, "ai_ledger": null, "ai_reading": "AI say-do: 2 GA (GPU nodes, 8-card plans), 1 beta without GA (App Platform); no operator assistant; MCP server is read-only and new." }, { "key": "civo", "name": "Civo Kubernetes (K3s based)", "short": "Civo Kubernetes", "kind": "product", "group": "developer", "since": 2021, "archetype": "Sovereign GPU cloud", "grade": "C", "driver": "M", "mix": { "V": 22, "C": 11, "M": 33, "U": 0, "P": 33 }, "idx": 0.33, "domains": { "lifecycle": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 2, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 2.5 }, "compute": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 2, "2022": 2, "2023": 2, "2024": 2.5, "2025": 2.5, "2026": 2.5 }, "network": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 2, "2022": 2, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 2.5 }, "storage": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 2, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 2 }, "security": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 2, "2025": 2, "2026": 2 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 2, "2026": 2 }, "fleet": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 1, "2022": 1, "2023": 1, "2024": 1, "2025": 1, "2026": 2 }, "devplat": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 2.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 3 }, "accel": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0, "2022": 0, "2023": 0, "2024": 2.5, "2025": 3, "2026": 3 }, "sovereignty": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 3, "2025": 3, "2026": 3 } }, "cred": [ [ "Cadence", 2, "No release-notes feed or version calendar; releases seen only via CNCF folders" ], [ "Roadmap transparency", 2, "Blog, newsroom and GitHub status page; no roadmap board" ], [ "Say-do", 3, "3 of 4 resolved product announcements shipped in 130 to 154 days; enclave Kubernetes alpha has no GA" ], [ "Velocity", 3, "GPU, Konstruct and sovereign-cloud products; few Kubernetes-service features in the record" ], [ "Lifecycle stability", 2, "No published support window; Flannel and Talos support being withdrawn" ] ], "comm": [ [ "Price and licence volatility", 0.5, "Hourly rates unchanged from 2024-02 to 2026; egress made free 2024-04-04" ], [ "Purchase constraints", 1, "Self-service pay as you go; GPUs also sold by commitment" ], [ "Channel and access", 1, "Public pricing and docs; five regions, GPU Kubernetes only in London" ], [ "Owner stability", 3.5, "Small private company with about GBP 1.4 million disclosed funding and a plan for 40 data centres" ], [ "Cost of staying supported", 2.5, "No support windows published; deprecations announced without dates" ] ], "risk": "commercial", "lock": [ [ "Formats", 1, "K3s is CNCF-certified Kubernetes; Talos and Flannel options being removed" ], [ "Export path", 0.5, "Free unlimited egress; no migration service found" ], [ "Stack coupling", 1.5, "civo-volume CSI and Civo load balancers and firewalls" ], [ "Hardware / cloud coupling", 2, "Civo cloud; CivoStack Enterprise and FlexCore run the same stack on own hardware" ], [ "Skills and tooling coupling", 1, "Civo CLI and K3s differences" ] ], "integ": [ [ "Partner programme openness", 2.5, "TD SYNNEX distributor 2025-04-23; agency partners" ], [ "Multi-tenancy and self-service", 2, "Accounts and Konstruct organisations; no fleet isolation" ], [ "API and infrastructure-as-code", 3.5, "Terraform provider 5.8 million downloads, CLI, API and Go client" ], [ "Skills and certification", 2.5, "Civo Academy and ambassadors; no Kubernetes certification path" ] ], "alt": null, "ranked": true, "reading": "Civo is moving from cheap K3s clusters toward a sovereign UK and India cloud with GPUs, a GitOps platform and private-cloud kits. The record shows fast new minors lately and free egress, but no release notes, no support policy and thin Kubernetes-service detail. A buyer gets simple pricing and a UK operator, with little documented lifecycle or enterprise depth and a small company behind a large build-out.", "report": "reports/12-civo.md", "ai": { "infra": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0, "2022": 0, "2023": 0, "2024": 1.5, "2025": 2, "2026": 2 }, "stack": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 1, "2026": 1 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 0 }, "agents": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 2 } }, "ai_basis": { "infra": "B200 and GH200 pools; GPU time-slicing explained in a blog", "stack": "relaxAI assistant and OpenAI-compatible API are separate Civo AI products", "ops": "None found", "agents": "MCP keys exist only in Konstruct; no Civo Kubernetes MCP server" }, "ai_ledger": null, "ai_reading": "AI say-do: B200 listed 2025-03-27 and live 2025-08-04, relaxAI unveiled 2024-09-11 and live 2025-02-11; no operator assistant; MCP only in Konstruct." }, { "key": "ovh", "name": "OVHcloud Managed Kubernetes Service (Free and Standard plans)", "short": "OVHcloud MKS", "kind": "product", "group": "european", "since": 2019, "archetype": "European cloud, two plans", "grade": "B", "driver": "C", "mix": { "V": 20, "C": 30, "M": 20, "U": 20, "P": 10 }, "idx": 0.29, "domains": { "lifecycle": { "2016": null, "2017": null, "2018": null, "2019": 2, "2020": 2, "2021": 2, "2022": 2, "2023": 2.5, "2024": 2.5, "2025": 3, "2026": 3 }, "compute": { "2016": null, "2017": null, "2018": null, "2019": 2, "2020": 2, "2021": 2.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 3, "2026": 3 }, "network": { "2016": null, "2017": null, "2018": null, "2019": 2, "2020": 2, "2021": 2, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 3, "2026": 3 }, "storage": { "2016": null, "2017": null, "2018": null, "2019": 2, "2020": 2, "2021": 2, "2022": 2, "2023": 2, "2024": 2.5, "2025": 2.5, "2026": 3 }, "security": { "2016": null, "2017": null, "2018": null, "2019": 1.5, "2020": 1.5, "2021": 2.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 3 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": 1, "2020": 1, "2021": 2, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 2 }, "fleet": { "2016": null, "2017": null, "2018": null, "2019": 1, "2020": 1, "2021": 1, "2022": 1, "2023": 1, "2024": 2, "2025": 2, "2026": 2 }, "devplat": { "2016": null, "2017": null, "2018": null, "2019": 1, "2020": 1, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 2 }, "accel": { "2016": null, "2017": null, "2018": null, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 2, "2024": 2.5, "2025": 2.5, "2026": 2.5 }, "sovereignty": { "2016": null, "2017": null, "2018": null, "2019": 1.5, "2020": 1.5, "2021": 2, "2022": 2, "2023": 2, "2024": 2, "2025": 3.5, "2026": 3.5 } }, "cred": [ [ "Cadence", 3, "Version issues on the roadmap; 1.36 missed its September 2026 promise" ], [ "Roadmap transparency", 4.5, "Public GitHub roadmap with answered issues and stated timings" ], [ "Say-do", 2.5, "4 of 9 resolved promises delivered within 12 months; Karpenter and Free-plan Cilium missed" ], [ "Velocity", 3, "Standard plan, Cilium, LUKS and custom CIDR in 2025 and 2026, but versions arrive slowly" ], [ "Lifecycle stability", 3, "Five minors offered, no extended support, third-oldest minor force-upgraded after 30 days" ] ], "comm": [ [ "Price and licence volatility", 2, "Paid Standard tier added 2025-09-17 and Cilium kept to it; Free plan unchanged" ], [ "Purchase constraints", 1, "Self-service pay as you go; Free plan" ], [ "Channel and access", 0.5, "Public docs, pricing and roadmap; 17 region codes" ], [ "Owner stability", 2, "Listed, family controlled; FY25 net income EUR 0.4 million and net debt EUR 1.1 billion" ], [ "Cost of staying supported", 2.5, "Support follows upstream; no extended support; forced upgrade of the third-oldest minor" ] ], "risk": "commercial", "lock": [ [ "Formats", 0.5, "CNCF-certified Kubernetes; standard APIs" ], [ "Export path", 1, "Free outbound traffic except Singapore and Sydney; migration and Velero guides" ], [ "Stack coupling", 1.5, "Octavia load balancers, Cinder volumes and vRack" ], [ "Hardware / cloud coupling", 3, "OVHcloud Public Cloud only; no hybrid extension found" ], [ "Skills and tooling coupling", 0.5, "kubectl, Terraform, Pulumi" ] ], "integ": [ [ "Partner programme openness", 2, "Terraform partner-tier provider; partner programme not verified" ], [ "Multi-tenancy and self-service", 2.5, "Public Cloud projects; Rancher Managed Service; no fleet isolation" ], [ "API and infrastructure-as-code", 4, "Terraform provider 39 million downloads, Pulumi, CDKTF, CLI and API" ], [ "Skills and certification", 2, "Documentation and tutorials; no Kubernetes certification path found" ] ], "alt": null, "ranked": true, "reading": "OVHcloud is building a paid Standard plan with 3-AZ control planes, Cilium and an SLA next to the free plan. The record shows a rare public roadmap with direct answers, but several dated promises slipped, Kubernetes 1.36 was late, and the free plan stays on Canal. A buyer gets a French operator, free outbound traffic and a clear roadmap, with slow version delivery and few managed add-ons.", "report": "reports/13-ovh.md", "ai": { "infra": { "2016": null, "2017": null, "2018": null, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 1.5 }, "stack": { "2016": null, "2017": null, "2018": null, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 0.5, "2024": 0.5, "2025": 0.5, "2026": 0.5 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": 0.5, "2020": 0.5, "2021": 0.5, "2022": 0.5, "2023": 0.5, "2024": 0.5, "2025": 0.5, "2026": 0.5 }, "agents": { "2016": null, "2017": null, "2018": null, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 2.5 } }, "ai_basis": { "infra": "A100 and A10 GPU flavors; guides for GPU apps and monitoring", "stack": "AI Endpoints and AI Deploy are separate products; no MKS integration found", "ops": "Docs search assistant only", "agents": "Labs MCP server with OAuth2 from 2026-02-27; Kubernetes tools not confirmed" }, "ai_ledger": null, "ai_reading": "AI say-do: 1 MCP request delivered as Labs in 107 days, 1 GPU-in-3-AZ request open; no assistant for operators and no AI service wired into MKS." }, { "key": "scaleway", "name": "Scaleway Kubernetes Kapsule and Kosmos", "short": "Scaleway Kapsule/Kosmos", "kind": "product", "group": "european", "since": 2019, "archetype": "European sovereign basics", "grade": "B", "driver": "C", "mix": { "V": 9, "C": 64, "M": 9, "U": 0, "P": 18 }, "idx": 0.11, "domains": { "lifecycle": { "2016": null, "2017": null, "2018": null, "2019": 0.5, "2020": 2, "2021": 2, "2022": 2, "2023": 2, "2024": 3, "2025": 3, "2026": 3 }, "compute": { "2016": null, "2017": null, "2018": null, "2019": 0.5, "2020": 2, "2021": 2, "2022": 2, "2023": 3, "2024": 3, "2025": 3, "2026": 3 }, "network": { "2016": null, "2017": null, "2018": null, "2019": 0.5, "2020": 2, "2021": 2, "2022": 2, "2023": 3, "2024": 3, "2025": 3, "2026": 3 }, "storage": { "2016": null, "2017": null, "2018": null, "2019": 0.5, "2020": 2, "2021": 2, "2022": 2, "2023": 2, "2024": 3, "2025": 3, "2026": 3.5 }, "security": { "2016": null, "2017": null, "2018": null, "2019": 0.5, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 3, "2025": 3, "2026": 3 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": 0.5, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 3, "2024": 3, "2025": 3, "2026": 3 }, "fleet": { "2016": null, "2017": null, "2018": null, "2019": 0.5, "2020": 0.5, "2021": 3.5, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "devplat": { "2016": null, "2017": null, "2018": null, "2019": 0.5, "2020": 1.5, "2021": 1.5, "2022": 3, "2023": 3, "2024": 3, "2025": 3, "2026": 3 }, "accel": { "2016": null, "2017": null, "2018": null, "2019": 1, "2020": 2, "2021": 2, "2022": 2, "2023": 3, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "sovereignty": { "2016": null, "2017": null, "2018": null, "2019": 1, "2020": 2, "2021": 2, "2022": 2, "2023": 3, "2024": 3, "2025": 3, "2026": 3.5 } }, "cred": [ [ "Cadence", 3.5, "Version calendar published and changelog dated, but 'a few days' availability claim missed; median lag 55 days" ], [ "Roadmap transparency", 3, "Dated changelog and status incident JSON; no public roadmap board; feature-request site could not be read" ], [ "Say-do", 3.5, "5 of 7 tracked items shipped on or before stated dates; version availability claim missed; Arm pools still beta" ], [ "Velocity", 3, "17 Kubernetes changelog entries in 2024 and 13 in 2025; about one substantive capability a quarter" ], [ "Lifecycle stability", 3, "14-month support, forced upgrades at end of support, several retirements (Flannel, Weave, public-only clusters, NAT IPs, CSI v0.2)" ] ], "comm": [ [ "Price and licence volatility", 1.5, "Free control plane capped at 150 nodes from 2023-09; dedicated tiers introduced; no list-price change from 2024-05 to 2026-10" ], [ "Purchase constraints", 1, "Self-service pay as you go; free shared control plane; 30-day commitment on dedicated planes" ], [ "Channel and access", 1.5, "Self-service sign-up and public pricing; four European regions only" ], [ "Owner stability", 1.5, "Iliad Group subsidiary; no distress found, but ownership and financials not examined" ], [ "Cost of staying supported", 2, "Patch upgrades free; minors supported 14 months, then forced upgrade within 30 days; no extended support" ] ], "risk": "commercial", "lock": [ [ "Formats", 0.5, "Certified Kubernetes (CNCF folders 1.24 to 1.37); standard API and kubeconfig" ], [ "Export path", 2, "No migration or Kubernetes backup service found; Object Storage egress 75 GB free then EUR 0.01 per GB" ], [ "Stack coupling", 2.5, "Scaleway IAM, Private Network, load balancers and block storage classes reconciled by the service; Private Network mandatory" ], [ "Hardware / cloud coupling", 2, "Kosmos adds nodes from other clouds, but the control plane runs only in Scaleway regions" ], [ "Skills and tooling coupling", 1, "kubectl, Terraform and a Scaleway CLI" ] ], "integ": [ [ "Partner programme openness", 3, "Partner programme with directory and application page; partner names not confirmed" ], [ "Multi-tenancy and self-service", 3.5, "Organisations, projects, IAM policies and applications" ], [ "API and infrastructure-as-code", 4, "Terraform provider 16.4 million downloads (partner-premier), CLI, Go, Python and JS SDKs, Crossplane provider; Cluster API provider is for self-managed clusters" ], [ "Skills and certification", 2.5, "Scaleway Learning page and community Slack; no Kubernetes certification found" ] ], "alt": null, "ranked": true, "reading": "Scaleway is building a plain, standard Kapsule service with a multi-cloud Kosmos option, French data location and GPU pools. The record shows steady dated changelog entries and kept deprecation timetables, but new minors arrive after a median 55 days against a stated few days, there is no roadmap board and no AI tooling beyond GPU pools. A buyer gets low cost, EU control and conformant Kubernetes, with an SLA only on paid dedicated control planes and four European regions.", "report": "reports/14-scaleway.md", "ai": { "infra": { "2016": null, "2017": null, "2018": null, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 1.5 }, "stack": { "2016": null, "2017": null, "2018": null, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 0 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 0.5, "2025": 0.5, "2026": 0.5 }, "agents": { "2016": null, "2017": null, "2018": null, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 1.5 } }, "ai_basis": { "infra": "GPU operator on every GPU pool 2023-07-18; scratch-storage guide for H100 and L40S pools; no sharing or inference add-ons", "stack": "none", "ops": "'Ask our AI' documentation search only; no cluster-aware assistant found", "agents": "REST API, CLI and Terraform; no official MCP server (community servers only)" }, "ai_ledger": null, "ai_reading": "AI say-do: 0 dated AI promises found; GPU pools with a default operator are shipped. Generative APIs and Managed Inference are separate products with no Kapsule integration; first-pass MCP server repository does not exist." }, { "key": "ionos", "name": "IONOS Cloud Managed Kubernetes", "short": "IONOS Managed Kubernetes", "kind": "product", "group": "european", "since": 2021, "archetype": "German low-cost basics", "grade": "B", "driver": "C", "mix": { "V": 0, "C": 44, "M": 33, "U": 11, "P": 11 }, "idx": 0.0, "domains": { "lifecycle": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 1.5, "2022": 1.5, "2023": 2, "2024": 2, "2025": 2.5, "2026": 2.5 }, "compute": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 2, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 3, "2026": 3 }, "network": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 1.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 2.5 }, "storage": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 1.5, "2022": 2, "2023": 2, "2024": 2, "2025": 2.5, "2026": 2.5 }, "security": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 1.5, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 2 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 1.5 }, "fleet": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 2, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 2 }, "devplat": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 1.5, "2022": 1.5, "2023": 2, "2024": 2, "2025": 2, "2026": 2 }, "accel": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 0 }, "sovereignty": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 2.5, "2025": 2.5, "2026": 2.5 } }, "cred": [ [ "Cadence", 3, "Release schedule with dates and monthly release notes, but new minors arrive 64 to 254 days after upstream and 1.36 is missing" ], [ "Roadmap transparency", 3, "Dated release notes, status incident updates and a vulnerability register; no roadmap; promised RCAs not found" ], [ "Say-do", 3, "5 items: 3 on time, 1 partial (NFS for Kubernetes still on request), 1 late (maintenance pause 64 days)" ], [ "Velocity", 2.5, "About 4 to 7 Kubernetes-specific release-note items a year; few new capabilities" ], [ "Lifecycle stability", 2.5, "End-of-life notices kept, but forced upgrades, a 64-day maintenance pause and 2026 control plane instability" ] ], "comm": [ [ "Price and licence volatility", 1, "No fee change found since the 2021 price list; control plane free; node types changed 2023 and 2025" ], [ "Purchase constraints", 1, "Self-service pay as you go with a 30-day trial; savings plans optional" ], [ "Channel and access", 1, "Self-service sign-up and public pricing; Kubernetes in four locations" ], [ "Owner stability", 1, "Listed company, United Internet holds 63.8%; revenue growing in 2026" ], [ "Cost of staying supported", 2.5, "Free upgrades but forced upgrade at end of life; product life about 11 to 14 months per minor" ] ], "risk": "commercial", "lock": [ [ "Formats", 0.5, "Certified Kubernetes (CNCF folders 1.27 to 1.35); Calico only" ], [ "Export path", 2.5, "Velero with CSI snapshots documented, no migration service; outbound traffic from USD 0.09 per GB on the US price list" ], [ "Stack coupling", 2.5, "IONOS block storage classes, contract-level permissions, LoadBalancer only on public pools; components reset at each maintenance" ], [ "Hardware / cloud coupling", 3.5, "IONOS Cloud only; no hybrid or on-premises extension found" ], [ "Skills and tooling coupling", 1, "kubectl, Terraform, DCD or ionosctl" ] ], "integ": [ [ "Partner programme openness", 3, "Channel partner programme and ISV programme; partner names not confirmed" ], [ "Multi-tenancy and self-service", 2.5, "Contracts with users, groups and privileges; no organisation or fleet isolation found" ], [ "API and infrastructure-as-code", 4, "Terraform provider 6.1 million downloads (partner-premier), Crossplane, Pulumi, Cluster API provider, SDKs and ionosctl" ], [ "Skills and certification", 2.5, "IONOS Cloud Learn material; no Kubernetes certification found" ] ], "alt": null, "ranked": true, "reading": "IONOS is building a plain, low-cost Managed Kubernetes on its own German cloud, adding an MCP server in 2026 but no GPU pools, fleet or AI features for clusters. The record shows slow version availability (median 119 days, 1.36 and 1.37 still missing), no control-plane SLA and a run of control plane outages in Frankfurt from March to September 2026. A buyer gets free control planes and German hosting, with one CNI, limited networking on private pools and weaker operations depth.", "report": "reports/15-ionos.md", "ai": { "infra": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 0 }, "stack": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 0 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 0 }, "agents": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 4 } }, "ai_basis": { "infra": "none", "stack": "none", "ops": "none", "agents": "Official MCP server GA 2026-06-09 with 8 Kubernetes read tools and opt-in write tools with confirmation; runs locally" }, "ai_ledger": null, "ai_reading": "AI say-do: 0 dated AI promises found. No GPU node pools or AI assistant for Kubernetes; GPU VMs and AI Model Hub are separate products; the MCP server is the only AI item." }, { "key": "vke", "name": "Vultr Kubernetes Engine (VKE)", "short": "Vultr VKE", "kind": "product", "group": "developer", "since": 2021, "archetype": "Low-cost GPU-ready basics", "grade": "B", "driver": "C", "mix": { "V": 0, "C": 44, "M": 33, "U": 11, "P": 11 }, "idx": 0.0, "domains": { "lifecycle": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0.5, "2022": 2, "2023": 2, "2024": 2.5, "2025": 2.5, "2026": 2.5 }, "compute": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0.5, "2022": 2.5, "2023": 2.5, "2024": 3.5, "2025": 3.5, "2026": 3.5 }, "network": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0.5, "2022": 2, "2023": 2, "2024": 3, "2025": 3, "2026": 3 }, "storage": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0.5, "2022": 2, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 2.5 }, "security": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 2.5, "2026": 2.5 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 1.5 }, "fleet": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0.5, "2022": 1.5, "2023": 1.5, "2024": 2, "2025": 2, "2026": 2 }, "devplat": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0.5, "2022": 1.5, "2023": 1.5, "2024": 2, "2025": 2, "2026": 2.5 }, "accel": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0, "2022": 0, "2023": 2, "2024": 3.5, "2025": 3.5, "2026": 4 }, "sovereignty": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0.5, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 2 } }, "cred": [ [ "Cadence", 3.5, "Every release dated in the changelog, 12 minors at 7 to 73 days; no published calendar" ], [ "Roadmap transparency", 2.5, "Dated changelog only; no roadmap, no support policy, status page keeps no history" ], [ "Say-do", 2.5, "Too few dated promises to judge: one pending MCP preview, one stale FAQ answer on bare metal" ], [ "Velocity", 3.5, "Bare metal, OIDC, VPC-only pools, snapshots, GPU pools and applications shipped 2024 to 2026" ], [ "Lifecycle stability", 2.5, "Manual upgrades that replace nodes and change IPs; no support window or end-of-life dates published" ] ], "comm": [ [ "Price and licence volatility", 1, "Control plane free since GA; pricing page text unchanged 2023 to 2026; HA fee not stated" ], [ "Purchase constraints", 0.5, "Self-service pay as you go, hourly billing" ], [ "Channel and access", 1, "Self-service sign-up, public docs and pricing, 33 locations" ], [ "Owner stability", 1.5, "Private company; financials not examined" ], [ "Cost of staying supported", 2.5, "No published support window; upgrades manual and replace nodes" ] ], "risk": "commercial", "lock": [ [ "Formats", 0.5, "Certified Kubernetes (CNCF folders 1.24 to 1.37); Calico only" ], [ "Export path", 2, "Cluster API and Terraform; no migration or Kubernetes backup service found; bandwidth overage fee not read" ], [ "Stack coupling", 2, "Vultr load balancers, block storage classes and VPC reconciled by the service; VPC 2.0 not supported" ], [ "Hardware / cloud coupling", 3.5, "Vultr cloud only; no hybrid or on-premises extension found" ], [ "Skills and tooling coupling", 1, "kubectl, Terraform and vultr-cli" ] ], "integ": [ [ "Partner programme openness", 2.5, "Partner programme and solution partners listed; tiers not read" ], [ "Multi-tenancy and self-service", 2.5, "Vultr IAM policies with Kubernetes actions; no organisation or fleet isolation read" ], [ "API and infrastructure-as-code", 4, "Terraform provider 6.1 million downloads (partner tier), Cluster API and Crossplane providers, CLI and Go client" ], [ "Skills and certification", 1.5, "No Kubernetes certification found" ] ], "alt": null, "ranked": true, "reading": "Vultr is building a simple, free-control-plane Kubernetes service that leans toward GPU and bare-metal capacity, with an MCP server in preview. The record shows fast version releases (median 41 days) and steady dated changelog entries, but no roadmap, no support policy, no control-plane SLA and no incident history. A buyer gets cheap standard Kubernetes with strong GPU choices and thin security, operations and sovereignty depth.", "report": "reports/16-vke.md", "ai": { "infra": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0, "2022": 0, "2023": 1, "2024": 2, "2025": 2, "2026": 2 }, "stack": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 1 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 0 }, "agents": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 3 } }, "ai_basis": { "infra": "GPU Operator and bare-metal GPU pools from 2024-05-14; no sharing or inference add-ons", "stack": "Applications tab deploys AMD AI Workbench and OpenFold3 (2026-06 and 2026-07); partner applications, maturity not stated", "ops": "none", "agents": "Official hosted MCP server in preview with OAuth, 11 Kubernetes tools, read-only, from 2026-05-14" }, "ai_ledger": null, "ai_reading": "AI say-do: 1 announced, 0 GA, 1 preview (MCP), 0 dropped. VKE has GPU pools and an applications catalogue but no AI assistant or inference add-on; first-pass GPU launch dates have no source." }, { "key": "ske", "name": "STACKIT Kubernetes Engine (SKE)", "short": "STACKIT SKE", "kind": "product", "group": "european", "since": 2020, "archetype": "Sovereign German managed K8s", "grade": "B", "driver": "C", "mix": { "V": 10, "C": 30, "M": 10, "U": 30, "P": 20 }, "idx": 0.2, "domains": { "lifecycle": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1, "2021": 1, "2022": 2, "2023": 2, "2024": 3, "2025": 3, "2026": 3 }, "compute": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1.5, "2021": 1.5, "2022": 2.5, "2023": 2.5, "2024": 3, "2025": 3, "2026": 3 }, "network": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1.5, "2021": 1.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 2.5 }, "storage": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1.5, "2021": 1.5, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 2.5 }, "security": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1, "2021": 1, "2022": 2, "2023": 2, "2024": 3, "2025": 3, "2026": 3.5 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1, "2021": 1, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 3 }, "fleet": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 2 }, "devplat": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1, "2021": 1, "2022": 1, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 2.5 }, "accel": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 2 }, "sovereignty": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 2.5, "2021": 2.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 4 } }, "cred": [ [ "Cadence", 2.5, "Patch days and version states published, but minors arrive 98 to 336 days after upstream against a stated 'shortly'" ], [ "Roadmap transparency", 3, "Dated release notes, lifecycle page, status incidents; no roadmap; promised RCA not found" ], [ "Say-do", 3, "Preview-to-supported windows kept; 'shortly after upstream' claim missed; audit logs, private clusters still in private preview" ], [ "Velocity", 3.5, "Workload Identity, IdP SSO, Gateway DNS, Ubuntu 26.04 and containerd 2 shipped 2025 to 2026" ], [ "Lifecycle stability", 3, "Clear deprecation notices and expiry dates, but about 8 to 11 months per minor and forced updates; no extended support" ] ], "comm": [ [ "Price and licence volatility", 2, "Cluster fee unchanged since at least 2025-12-12 but earlier prices not found; SKU split for Confidential Kubernetes 2025-12-01; larger node reservations 2025-07" ], [ "Purchase constraints", 1.5, "Pay as you go by the hour with hibernation; contract-based portal onboarding not confirmed" ], [ "Channel and access", 2, "Public docs and price table; two regions; self-service sign-up terms not confirmed" ], [ "Owner stability", 0.5, "Schwarz Group, a large privately held retailer" ], [ "Cost of staying supported", 3, "Minors expire about 8 to 11 months after release and are force-updated; no extended support" ] ], "risk": "commercial", "lock": [ [ "Formats", 1, "Certified Kubernetes (folder stackit-kubernetes-engine, v1.24 to v1.36) built on open-source Gardener; STACKIT IdP, ACL and Workload Identity extensions" ], [ "Export path", 1.5, "Service certificate lists exportable data; Velero how-to; no migration service found; egress price not read" ], [ "Stack coupling", 2.5, "STACKIT block storage classes, network load balancers, project IAM and IdP, SNA networks" ], [ "Hardware / cloud coupling", 3, "Runs only on STACKIT, though providers and extensions are open source" ], [ "Skills and tooling coupling", 1, "kubectl, Terraform and the STACKIT CLI" ] ], "integ": [ [ "Partner programme openness", 3, "Marketplace partner programme; tiers not read" ], [ "Multi-tenancy and self-service", 3.5, "Projects, Network Areas and a Landing Zone Accelerator for organisations" ], [ "API and infrastructure-as-code", 4.5, "Terraform provider 19.2 million downloads, Pulumi, CLI, SDKs for Go, Python and Java, Cluster API provider" ], [ "Skills and certification", 2, "No Kubernetes certification found" ] ], "alt": null, "ranked": true, "reading": "STACKIT is building a sovereign German managed Kubernetes on open-source Gardener, adding identity, Gateway API and Ubuntu support while private clusters, audit logs and Cilium stay in private preview. The record shows strong compliance scope (C5 Type 2, ISO 27001, TISAX) and clear lifecycle rules, but new Kubernetes minors arrive after a median of at least 143 days, minors expire after 8 to 11 months and the control plane is charged per cluster. A buyer gets German hosting and certification at the cost of slow version uptake and few GPU or fleet features.", "report": "reports/17-ske.md", "ai": { "infra": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 1 }, "stack": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 0 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 0 }, "agents": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 1.5 } }, "ai_basis": { "infra": "GPU flavors with a customer-installed operator; Ubuntu 26.04 GPU support 2026-10-06", "stack": "none", "ops": "none", "agents": "REST API, CLI and Terraform; no official MCP server" }, "ai_ledger": null, "ai_reading": "AI say-do: 0 dated AI promises found. SKE offers GPU flavors but no inference add-on or assistant; AI Model Serving is a separate product." }, { "key": "coreweave", "name": "CoreWeave Kubernetes Service (CKS)", "short": "CoreWeave CKS", "kind": "product", "group": "aicloud", "since": 2025, "archetype": "Sales-gated GPU cloud", "grade": "B", "driver": "C", "mix": { "V": 10, "C": 50, "M": 40, "U": 0, "P": 0 }, "idx": 0.1, "domains": { "lifecycle": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": null, "2022": null, "2023": null, "2024": null, "2025": 2.5, "2026": 3.5 }, "compute": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": null, "2022": null, "2023": null, "2024": null, "2025": 3, "2026": 3.5 }, "network": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": null, "2022": null, "2023": null, "2024": null, "2025": 3, "2026": 3.5 }, "storage": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": null, "2022": null, "2023": null, "2024": null, "2025": 3, "2026": 3.5 }, "security": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": null, "2022": null, "2023": null, "2024": null, "2025": 3, "2026": 3.5 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": null, "2022": null, "2023": null, "2024": null, "2025": 3, "2026": 3.5 }, "fleet": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": null, "2022": null, "2023": null, "2024": null, "2025": 2, "2026": 2 }, "devplat": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": null, "2022": null, "2023": null, "2024": null, "2025": 2.5, "2026": 3 }, "accel": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": null, "2022": null, "2023": null, "2024": null, "2025": 4, "2026": 4.5 }, "sovereignty": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": null, "2022": null, "2023": null, "2024": null, "2025": 2, "2026": 2 } }, "cred": [ [ "Cadence", 4, "Three minors a year, last four supported within 7 to 36 days; dated changelog; one early minor took 184 days" ], [ "Roadmap transparency", 3, "Detailed dated release notes and docs, but no public roadmap board and no stated delivery dates" ], [ "Say-do", 3, "2 of 2 completed preview-to-GA moves took 11 and 29 days; 6 items pending, none dated; evidence thin" ], [ "Velocity", 5, "Several substantive capabilities every quarter: GPU generations, autoscaler, spot, IAM, prefill, MCP" ], [ "Lifecycle stability", 3, "Three latest minors only; older versions deprecated for new clusters; upgrades one minor at a time; no extended support" ] ], "comm": [ [ "Price and licence volatility", 1.5, "H100 price unchanged from 2025-01-01 to 2026-10-13; storage tiers 2025-10 and quota cuts 2026-06" ], [ "Purchase constraints", 4, "Committed take-or-pay contracts were over 98% of 2025 revenue; on-demand exists but account approval needs Sales" ], [ "Channel and access", 4.5, "Sales approves every organisation; pricing page shows Contact Sales; docs are public" ], [ "Owner stability", 4, "Net loss $1.17 billion in 2025, debt $21.6 billion, Microsoft 67% of revenue; public company since 2025" ], [ "Cost of staying supported", 2.5, "Control plane free and no extended-support fee, but only three minors supported and upgrades are sequential" ] ], "risk": "commercial", "lock": [ [ "Formats", 1.5, "Conformant Kubernetes, but nodes come from NodePool custom resources and ncore images" ], [ "Export path", 1, "No ingress, egress or transfer fees and S3-compatible storage; no migration or backup service" ], [ "Stack coupling", 3, "InfiniBand, DPU VPCs, LOTA object storage, Distributed File Storage and SUNK are CoreWeave-specific" ], [ "Hardware / cloud coupling", 4, "Runs only on CoreWeave bare metal; no hybrid or other-cloud extension" ], [ "Skills and tooling coupling", 1.5, "Standard kubectl and Terraform plus CoreWeave NodePool and API specifics" ] ], "integ": [ [ "Partner programme openness", 2.5, "Partner programme page not opened; CNCF member since 2025-09; AI Conformance submissions" ], [ "Multi-tenancy and self-service", 4, "Organisations, IAM Access Policies, SCIM groups, per-zone quotas and single-tenant zones" ], [ "API and infrastructure-as-code", 4, "CKS API GA and Terraform provider with 552,397 downloads; no Cluster API provider found" ], [ "Skills and certification", 2.5, "No CoreWeave certification found; standard Kubernetes skills apply" ] ], "alt": null, "ranked": false, "reading": "CoreWeave is building toward a bare-metal AI cloud where Kubernetes, InfiniBand, storage and agent tooling come as one stack. The record shows fast Kubernetes tracking since 2025, a steady stream of GPU generations and IAM work, and no stated dates. A buyer gets top GPU access and free egress, and accepts sales-gated onboarding, committed contracts and a loss-making, debt-heavy supplier.", "report": "reports/18-coreweave.md", "ai": { "infra": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": null, "2022": null, "2023": null, "2024": null, "2025": 4, "2026": 4.5 }, "stack": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": null, "2022": null, "2023": null, "2024": null, "2025": 2, "2026": 3.5 }, "ops": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": null, "2022": null, "2023": null, "2024": null, "2025": 0, "2026": 1.5 }, "agents": { "2016": null, "2017": null, "2018": null, "2019": null, "2020": null, "2021": null, "2022": null, "2023": null, "2024": null, "2025": 2, "2026": 4.5 } }, "ai_basis": { "infra": "Rack-level pools, IMEX with DRA in limited availability, B300 and Vera Rubin", "stack": "CoreWeave Inference available 2026-03-31, with an on-CKS mode", "ops": "Mission Control Agent public preview 2026-09-24: answers with evidence, limited approved writes", "agents": "Mission Control MCP GA 2026-09-24 with token-scoped access and per-organisation write tools" }, "ai_ledger": null, "ai_reading": "AI say-do: 5 tracked, 1 GA (MCP), 1 partial (Agent preview), 3 pending (IMEX DRA, Skills, Vera Rubin). The Agent and Skills are previews; the MCP server is the one GA agent feature." }, { "key": "otc", "name": "Open Telekom Cloud Cloud Container Engine (CCE, CCE Turbo), now T Cloud Public", "short": "Open Telekom Cloud CCE", "kind": "product", "group": "european", "since": 2018, "archetype": "Sovereign German CCE", "grade": "B", "driver": "P", "mix": { "V": 10, "C": 20, "M": 10, "U": 0, "P": 60 }, "idx": 0.25, "domains": { "lifecycle": { "2016": null, "2017": null, "2018": 2, "2019": 2, "2020": 2, "2021": 2, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 3 }, "compute": { "2016": null, "2017": null, "2018": 1.5, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 3 }, "network": { "2016": null, "2017": null, "2018": 1.5, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 3 }, "storage": { "2016": null, "2017": null, "2018": 1.5, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 3 }, "security": { "2016": null, "2017": null, "2018": 2, "2019": 2, "2020": 2, "2021": 2, "2022": 3, "2023": 3, "2024": 3, "2025": 3, "2026": 3.5 }, "ops": { "2016": null, "2017": null, "2018": 1, "2019": 1, "2020": 1, "2021": 1, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 3 }, "fleet": { "2016": null, "2017": null, "2018": 1, "2019": 1, "2020": 1, "2021": 1, "2022": 1, "2023": 1, "2024": 1, "2025": 1, "2026": 2 }, "devplat": { "2016": null, "2017": null, "2018": 1.5, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 3 }, "accel": { "2016": null, "2017": null, "2018": 1, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 3.5 }, "sovereignty": { "2016": null, "2017": null, "2018": 3, "2019": 3, "2020": 3, "2021": 3, "2022": 3.5, "2023": 3.5, "2024": 3.5, "2025": 3.5, "2026": 4 } }, "cred": [ [ "Cadence", 3, "Minor release about every 4 to 5 months since 2025, but each about 10 months behind upstream and no published calendar" ], [ "Roadmap transparency", 4, "Public roadmap page with archived history plus dated release notes and a documented version policy" ], [ "Say-do", 3.5, "6 of 8 roadmap items shipped within 12 months of first listing; CCE Turbo GA took 18.5 months, 1.33 took 15" ], [ "Velocity", 3, "One or two substantive capabilities a quarter: Turbo GA, new minors, H100, add-ons" ], [ "Lifecycle stability", 2.5, "About two years per version, but end-of-service dates were shortened by 3 to 9 months in 2025 and 1.26 was skipped" ] ], "comm": [ [ "Price and licence volatility", 2, "Common I/O disk end of sales 2025-07-01 and decommissioning 2026-12-31, GPU family retirements, rename to T Cloud Public; no control-plane fee change found" ], [ "Purchase constraints", 1, "Self-service booking with a 250 euro credit; pay-as-you-go" ], [ "Channel and access", 1, "Public pricing models and release notes; docs site behind a bot challenge; two regions" ], [ "Owner stability", 1.5, "Deutsche Telekom subsidiary; reorganised into the T Cloud division in 2025 and renamed in 2026" ], [ "Cost of staying supported", 2.5, "Two-year support windows without fees, but versions arrive about 10 months late and end-of-service dates were cut" ] ], "risk": "commercial", "lock": [ [ "Formats", 1.5, "Kubernetes API with Everest storage classes and ELB annotations; no CNCF conformance listing found" ], [ "Export path", 2, "Velero-based migration guide; egress charges not read" ], [ "Stack coupling", 2.5, "Everest CSI, ELB ingress and Turbo pod ENIs with provider security groups" ], [ "Hardware / cloud coupling", 4, "Runs only in the two Open Telekom Cloud regions; UCS hybrid is planned, not released" ], [ "Skills and tooling coupling", 1.5, "Standard kubectl and Terraform plus CCE console and add-on specifics" ] ], "integ": [ [ "Partner programme openness", 3.5, "CIRCLE and TECH partner programmes listed on the site; not opened" ], [ "Multi-tenancy and self-service", 3.5, "IAM projects and enterprise dashboards; per-region cluster quotas" ], [ "API and infrastructure-as-code", 4.5, "Terraform provider with 7.2 million downloads, Crossplane provider 1.0 on 2026-09-04, Python SDK" ], [ "Skills and certification", 3.5, "T Cloud Public certification programme since 2020-03-17 and a training catalogue" ] ], "alt": null, "ranked": false, "reading": "T Cloud Public builds toward a German sovereign container service on Huawei-derived CCE technology, with Turbo clusters, H100 nodes and add-ons as the next steps. The record shows a public roadmap that is mostly kept but late, Kubernetes minors that arrive about ten months after upstream, and end-of-service dates that were shortened in 2025. A buyer gets German operation, C5 and TISAX, and pays with old Kubernetes versions and no CNCF conformance listing.", "report": "reports/19-otc.md", "ai": { "infra": { "2016": null, "2017": null, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 1.5, "2024": 1.5, "2025": 3, "2026": 3 }, "stack": { "2016": null, "2017": null, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 1, "2024": 1, "2025": 1, "2026": 1.5 }, "ops": { "2016": null, "2017": null, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 1 }, "agents": { "2016": null, "2017": null, "2018": 1.5, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 1.5 } }, "ai_basis": { "infra": "New H100 instance 2026-05-29; NVLink H100 only on the roadmap", "stack": "ModelArts updates through 2026-09; Volcano scheduler on CCE", "ops": "A support chatbot is listed; its scope was not read", "agents": "REST API, Terraform provider and Python SDK; no MCP server found" }, "ai_ledger": null, "ai_reading": "AI say-do: 3 tracked, 1 shipped (H100 in CCE), 2 pending (NVLink H100, confidential containers). GPU nodes are the substance; no MCP server and no confirmed assistant." }, { "key": "upstream", "name": "Upstream Kubernetes (kubernetes/kubernetes, CNCF)", "short": "Upstream Kubernetes", "kind": "product", "group": "engine", "since": 2016, "archetype": "Reference engine, DIY", "grade": "A", "driver": "C", "mix": { "V": 30, "C": 50, "M": 20, "U": 0, "P": 0 }, "idx": 0.3, "domains": { "lifecycle": { "2016": 2, "2017": 2, "2018": 2, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 2.5 }, "compute": { "2016": 1, "2017": 1, "2018": 1, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 2 }, "network": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 2, "2020": 2, "2021": 2, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 3 }, "storage": { "2016": 1, "2017": 1, "2018": 1, "2019": 2, "2020": 2, "2021": 2, "2022": 2.5, "2023": 2.5, "2024": 2.5, "2025": 2.5, "2026": 3 }, "security": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 2.5, "2020": 2.5, "2021": 2.5, "2022": 3, "2023": 3, "2024": 3, "2025": 3, "2026": 3.5 }, "ops": { "2016": 1, "2017": 1, "2018": 1, "2019": 1, "2020": 1, "2021": 1, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 2 }, "fleet": { "2016": 0.5, "2017": 0.5, "2018": 0.5, "2019": 1, "2020": 1, "2021": 1, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 1.5 }, "devplat": { "2016": 1, "2017": 1, "2018": 1, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 2 }, "accel": { "2016": 0.5, "2017": 0.5, "2018": 0.5, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 2, "2023": 2, "2024": 2, "2025": 2, "2026": 3.5 }, "sovereignty": { "2016": 0, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 0 } }, "cred": [ [ "Cadence", 5, "Three minors a year; 10 of 11 releases from 1.27 to 1.37 shipped on the date published at cycle start" ], [ "Roadmap transparency", 5, "Public KEP repository with commit history, sig-release schedules and an official CVE feed" ], [ "Say-do", 4, "20 of 24 stated targets met within 12 months; 16 of 24 on the stated date; sidecars, resize, swap and group snapshots slipped over a year" ], [ "Velocity", 5, "11 to 23 features reach stable in every release; 44 to 70 enhancements per release since 1.27" ], [ "Lifecycle stability", 2.5, "About 14 months per minor, no long-term support release; upgrade every minor, APIs get removed" ] ], "comm": null, "risk": "engine", "lock": [ [ "Formats", 0, "The reference API; workloads use standard Kubernetes objects" ], [ "Export path", 0, "Nothing to export: runs anywhere, no egress fee, no vendor service" ], [ "Stack coupling", 0, "No provider IAM, load balancer or storage class required" ], [ "Hardware / cloud coupling", 0, "Runs on any cloud, bare metal or laptop" ], [ "Skills and tooling coupling", 0.5, "Standard kubectl; operating etcd and kubeadm yourself needs scarcer skills" ] ], "integ": [ [ "Partner programme openness", 4.5, "Vendor-neutral CNCF conformance programme with 131 listed products for 1.34 and 1.35" ], [ "Multi-tenancy and self-service", 2.5, "Namespaces, RBAC and quotas only; no organisations, projects or fleet isolation" ], [ "API and infrastructure-as-code", 4, "Cluster API v1.14 and client libraries; no official Terraform provider" ], [ "Skills and certification", 5, "CKA exam offered by the Linux Foundation and the largest public documentation and community" ] ], "alt": null, "ranked": true, "reading": "Upstream Kubernetes is building toward a standard scheduling and admission layer for accelerators and policy, with a release train that holds its dates. The record shows reliable releases and features that often take two to five extra minors to reach stable. A buyer gets full portability and no fee, and pays with operations: no hosted control plane, no node provisioning, 14 months of support.", "report": "reports/20-upstream.md", "ai": { "infra": { "2016": 0, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 1, "2024": 1, "2025": 3, "2026": 3.5 }, "stack": { "2016": 0, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 0 }, "ops": { "2016": 0, "2017": 0, "2018": 0, "2019": 0, "2020": 0, "2021": 0, "2022": 0, "2023": 0, "2024": 0, "2025": 0, "2026": 0 }, "agents": { "2016": 1.5, "2017": 1.5, "2018": 1.5, "2019": 1.5, "2020": 1.5, "2021": 1.5, "2022": 1.5, "2023": 1.5, "2024": 1.5, "2025": 1.5, "2026": 1.5 } }, "ai_basis": { "infra": "DRA extended resources and device taints stable 1.37; agent-sandbox 1.0.0 on 2026-08-28", "stack": "none", "ops": "The project ships no assistant; k8sgpt and kubectl-ai are third-party", "agents": "Documented REST API; no official MCP server (the draft's kubernetes-sigs repository does not exist)" }, "ai_ledger": null, "ai_reading": "AI say-do: 3 tracked, 2 GA, 1 pending (DRA consumable capacity slipped from 1.36 to 1.38). The project supplies scheduling primitives for GPUs; it has no assistant and no official MCP server." } ] }