UNOFFICIAL/EXPERIMENTAL. This entry belongs to the AICR catalog overview. It was produced by scripts/generate-aicr-from-overlay.mjs l40s-oke-inference, the generator built to mirror any overlay from NVIDIA's AICR recipe catalog into a first-class entry at the same bar as this catalog's other entries: source pinned by digest, render checked, one digest over the whole rendered shape.
What it is
aicr recipe --service oke --accelerator l40s --os ol --intent inference --output recipe.yaml
aicr bundle --recipe recipe.yaml --deployer argocd-helm --output ./argocd-helm-bundle \
--storage-class gp3 --accelerated-node-selector nvidia.com/gpu.present=true \
--workload-selector app.kubernetes.io/part-of=inference
helm template aicr-argocd ./argocd-helm-bundle --namespace argocd \
--set repoURL=oci://europe-west1-docker.pkg.dev/nth-fort-499605-q5/helm-expt --output-dir rendered
The generation receipt records the exact commands, the criteria, and the release binary this entry was built with: AICR v0.21.0, verified against the checksum published in that release before it ran.
One digest pins this entry too
node scripts/generate-aicr-digest-index.mjs --verify --example l40s-oke-inference
The digest index pins the upstream source, the recipe criteria, and all 16 Applications under one platform digest. The compiler is a generic one shared by every entry in this catalog, not something built new for this one.
What is proven and what is not
Proven: the entry was generated by a binary verified against a checksum pinned before it ran, the recipe and every rendered Application are retained byte for byte, and one digest pins the shape.
Not proven, and stated rather than implied: this entry was never published, so it carries no public digest and no OCI transport receipts, and the repository reference in its Applications is what a publication would use rather than a record that one happened. No ConfigHub import or promotion exists for it, no cluster ever ran it, and no NGC surface was contacted.