NVIDIA/aicr

View on GitHub

Tooling for optimized, validated, and reproducible GPU-accelerated AI runtime in Kubernetes

Stars: 422
Forks: 101
Open beginner issues: 3
Indexed issues: 173
Avg merge: 1d 1h
Merged PRs (30d): 267

Dominant language: Go
License: Apache-2.0
Last GitHub push: Sep 20, 2026
Latest indexed: Sep 19, 2026
Contributing guide: Contributing guide
Code of conduct: Code of conduct
Beginner labels: No beginner labels indexed

Beginner friendly Everything

Open All

168 open issues indexed

recipe list: JSON and YAML disagree on criteria key casing Open

area/cli area/sdk

Difficulty 2/5 1-3 hours Newbie friendliness 86/100

NVIDIA/aicr#2897 · Opened Sep 21, 2026
bundle --attest signs before verifying the binary attestation digest Open

area/bundler theme/supply-chain

Difficulty 3/5 1-2 days Newbie friendliness 74/100

NVIDIA/aicr#2896 · Opened Sep 21, 2026
bundle: --image-refs rejects any path under a symlinked parent Open

area/bundler area/cli

Difficulty 3/5 1-2 days Newbie friendliness 78/100

NVIDIA/aicr#2895 · Opened Sep 21, 2026
createCommitOnBranch callers die on jq ARG_MAX for large files (render-golden-refresh, evidence) Open

area/ci theme/ci-dx

NVIDIA/aicr#2893 · 1 assignee · Opened Sep 21, 2026
Retire two superseded GB300 evidence pointers naming :edge validators Open

area/validator theme/supply-chain

NVIDIA/aicr#2883 · 1 assignee · Opened Sep 20, 2026
Record resolved validator image digests in recipe-evidence predicate Open

area/validator theme/supply-chain

NVIDIA/aicr#2882 · 1 assignee · Opened Sep 20, 2026
Render evidence trust class in the recipe-health Evidence column Open

area/docs theme/validation

Difficulty 3/5 1-2 days Newbie friendliness 58/100

NVIDIA/aicr#2877 · Opened Sep 20, 2026
Refresh VR200 RKE2 evidence at v1.0.0-rc1, after the nvsentinel bump Open

area/docs area/recipes theme/validation

NVIDIA/aicr#2876 · 1 assignee · Opened Sep 20, 2026
Publish GB300 EKS evidence from a v1.0.0-rc1 run Open

area/docs area/recipes theme/validation

NVIDIA/aicr#2875 · 1 assignee · Opened Sep 20, 2026
EKS: nvidia-setup-kernel should no-op before apt when the running kernel already matches Open

area/recipes theme/recipes

NVIDIA/aicr#2870 · 4 comments · 2 assignees · Opened Sep 19, 2026
Flaky: nvsentinel-object-monitor/health-event races the event it never waits for Open

area/tests theme/ci-dx

NVIDIA/aicr#2866 · 3 comments · 1 assignee · Opened Sep 19, 2026
EKS: verify whether nvidia-setup-kernel's 4 CPU / 8Gi suffices for a genuine kernel install Open

area/recipes theme/recipes

NVIDIA/aicr#2865 · 4 comments · 2 assignees · Opened Sep 19, 2026
[Bug]: readiness gate is broken with --deployer helm due to missing hook annotations Open

area/cli

NVIDIA/aicr#2862 · 2 comments · 1 assignee · Opened Sep 19, 2026
UAT release cells run main's test fixtures against a released binary Open

area/ci area/tests theme/ci-dx

NVIDIA/aicr#2860 · 1 assignee · Opened Sep 19, 2026
[Feature]: Terraform deployer Open

area/cli

Difficulty 5/5 Over a week Newbie friendliness 25/100

NVIDIA/aicr#2857 · 2 comments · Opened Sep 19, 2026
CRD step error message prints the connection flag's value and misstates the rejection reason Open

area/bundler theme/supply-chain

Difficulty 2/5 1-3 hours Newbie friendliness 88/100

NVIDIA/aicr#2854 · 1 comment · 1 assignee · Opened Sep 19, 2026
dra-node-labeler health check suppressed on the DRA-eviction opt-in path Open

area/recipes area/validator theme/validation

NVIDIA/aicr#2848 · 1 assignee · Opened Sep 19, 2026
fix(bundler): unquoted name/namespace in localformat install.sh templates Open

area/bundler theme/deployer

Difficulty 3/5 1-2 days Newbie friendliness 82/100

NVIDIA/aicr#2844 · 1 comment · Opened Sep 18, 2026
Fix the documented CRD fallback for existing resources Open

area/docs theme/deployer

NVIDIA/aicr#2843 · 1 assignee · Opened Sep 18, 2026
Docs claim a Helmfile CRD automation that does not exist Open

area/docs theme/deployer

NVIDIA/aicr#2841 · 1 assignee · Opened Sep 18, 2026
docs: link gate checks file existence, not published reachability Open

area/ci area/docs theme/ci-dx

Difficulty 3/5 1-2 days Newbie friendliness 75/100

NVIDIA/aicr#2838 · Opened Sep 18, 2026
[Feature]: Add a GPUDirect-RDMA-wired torch-distributed-rdma ClusterTrainingRuntime for the GB200 GKE Kubeflow leaf (parity with H100/TCPXO) Open

area/recipes

NVIDIA/aicr#2837 · 1 assignee · Opened Sep 18, 2026
validator: nodewright readiness can skip on a CRD-establishment race Open

area/validator theme/validation

Difficulty 4/5 3-5 days Newbie friendliness 62/100

NVIDIA/aicr#2836 · 1 comment · Opened Sep 18, 2026
Component object names pinned by fullnameOverride are unguarded Open

area/bundler theme/recipes

NVIDIA/aicr#2835 · 1 comment · 1 assignee · Opened Sep 18, 2026
Guard the other five registry-derived identity fields, not just namespace Open

area/bundler theme/recipes

NVIDIA/aicr#2834 · 1 comment · 1 assignee · Opened Sep 18, 2026
[Bug]: Embedded and Layered DataProvider Fails on Nested Paths Across Non-POSIX Filesystems Open

area/recipes

Difficulty 3/5 1-2 days Newbie friendliness 84/100

NVIDIA/aicr#2831 · 2 comments · Opened Sep 18, 2026
Pin deployment identity at first deploy so registry defaults can move safely Open

area/bundler theme/recipes

NVIDIA/aicr#2830 · 1 comment · 1 assignee · Opened Sep 18, 2026
Wire dra-node-labeler on OpenShift (gpu-operator-ocp / nvidia-dra-driver-gpu-ocp) Open

area/bundler area/cli area/recipes enhancement theme/recipes

Difficulty 4/5 3-5 days Newbie friendliness 45/100

NVIDIA/aicr#2828 · Opened Sep 18, 2026
v1.0.0 release execution: correct the v0.23 binding, freeze the baselines, cut the tag Open

area/ci area/docs theme/ci-dx

NVIDIA/aicr#2807 · 1 assignee · Opened Sep 17, 2026
topograph repo move: dead chart index, and unpublished image at the new package Open

area/recipes theme/recipes

NVIDIA/aicr#2793 · 2 comments · 1 assignee · Opened Sep 17, 2026
enforce kube-prometheus-stack / prometheus-operator-crds appVersion lockstep Open

area/recipes theme/recipes

Difficulty 4/5 3-5 days Newbie friendliness 64/100

NVIDIA/aicr#2792 · Opened Sep 17, 2026
drift report proposes stale and retracted chart versions (max-semver selection) Open

area/ci theme/ci-dx

NVIDIA/aicr#2791 · 1 assignee · Opened Sep 17, 2026
Scheduled maintenance workflows emit no signal when the run itself fails Open

area/ci theme/ci-dx

NVIDIA/aicr#2788 · 1 assignee · Opened Sep 16, 2026
Use Beacon baselines as a reference point for NCCL performance gates Open

area/recipes area/validator theme/validation

Difficulty 5/5 Over a week Newbie friendliness 35/100

NVIDIA/aicr#2787 · Opened Sep 16, 2026
ci(uat): GCP janitor never discovers orphans whose cluster and VPC are already gone Open

area/ci theme/ci-dx

NVIDIA/aicr#2775 · 1 assignee · Opened Sep 16, 2026
EBS provisioning readiness with actionable IAM/auth diagnostics Open

area/validator

Difficulty 5/5 Over a week Newbie friendliness 38/100

NVIDIA/aicr#2766 · 2 comments · Opened Sep 15, 2026
First-class self-managed-on-AWS H100/Dynamo recipe coverage Open

area/recipes

Difficulty 4/5 3-5 days Newbie friendliness 48/100

NVIDIA/aicr#2765 · 2 comments · Opened Sep 15, 2026
Stop stock_render_golden.yaml conflicting on every bundler rebase Open

area/bundler theme/ci-dx

Difficulty 4/5 3-5 days Newbie friendliness 48/100

NVIDIA/aicr#2761 · 2 comments · Opened Sep 15, 2026
No check verifies GPU access or isolation on Slinky Slurm recipes Open

area/validator theme/validation

Difficulty 4/5 3-5 days Newbie friendliness 55/100

NVIDIA/aicr#2756 · 1 comment · Opened Sep 14, 2026
x86 OKE/LKE/RTX PRO 6000 training bases omit robust-controller and secure-accelerator-access Open

area/recipes theme/validation

Difficulty 4/5 3-5 days Newbie friendliness 52/100

NVIDIA/aicr#2718 · 1 comment · Opened Sep 11, 2026
Track the upstream gpu-operator fix for the computedomains CRD divergence Open

area/bundler theme/deployer

NVIDIA/aicr#2715 · 1 comment · 1 assignee · Opened Sep 11, 2026
[Feature]: Measure CRE checks on non-H100 hardware without changing shipped recipes Open

area/validator theme/validation

NVIDIA/aicr#2714 · 1 comment · 1 assignee · Opened Sep 11, 2026
[Feature]: Generate a performance-coverage report with Qualified/Experimental/Blocked/Unsupported states Open

area/recipes theme/validation

NVIDIA/aicr#2701 · 1 comment · 1 assignee · Opened Sep 10, 2026
[Feature]: Attach calibrated CRE constraints to the seven NCCL overlays Open

area/recipes theme/validation

NVIDIA/aicr#2700 · 1 comment · 1 assignee · Opened Sep 10, 2026
[Feature]: Upstream - fix H200 EFA selection and add oci/forge catalog entries Open

area/validator theme/validation

NVIDIA/aicr#2699 · 1 comment · 1 assignee · Opened Sep 10, 2026
[Feature]: Upstream - publish the NVCRE workload runtime closure as pinned artifacts Open

area/recipes theme/supply-chain

NVIDIA/aicr#2698 · 1 comment · 1 assignee · Opened Sep 10, 2026
[Feature]: Upstream - bound Certification node footprint and total run duration Open

area/validator theme/validation

NVIDIA/aicr#2697 · 1 comment · 1 assignee · Opened Sep 10, 2026
[Feature]: Fail make bom-docs loudly when a chart pull returns no images Open

area/ci theme/ci-dx

NVIDIA/aicr#2696 · 1 comment · 1 assignee · Opened Sep 10, 2026
[Feature]: Enforce parity between registry namespaces and tools/cleanup Open

area/infra theme/ci-dx

NVIDIA/aicr#2695 · 2 comments · 1 assignee · Opened Sep 10, 2026
[Feature]: Add an adaptive fault isolation check driving CRE bisection Open

area/validator theme/validation

NVIDIA/aicr#2693 · 1 comment · 1 assignee · Opened Sep 10, 2026

NVIDIA/aicr

Get new issues in your inbox