Evan Lezar 294c64eed1 fix(gpu): prefer single CDI devices for local runtimes (#1675) ...

Prefer a single CDI-qualified device when Docker or Podman resolves the default GPU request to one GPU.

Allow nvidia.com/gpu=all only as a WSL2 all-only compatibility fallback, using Docker daemon info and Podman's /dev/dxg probe to identify that case.

Update driver docs, architecture notes, and GPU e2e coverage for the default selection behavior.

Signed-off-by: Evan Lezar <elezar@nvidia.com>

2026-06-16 10:14:49 +02:00

..

inference-routing.mdx

feat(inference): allow local embeddings route (#1774)

2026-06-07 10:27:34 -07:00

manage-gateways.mdx

feat(gateway): add system registry support and source indicators (#1625)

2026-06-11 16:25:12 -07:00

manage-providers.mdx

feat(cli): add JSON/YAML output format to provider list command (#1830)

2026-06-15 14:52:50 -07:00

manage-sandboxes.mdx

fix(gpu): prefer single CDI devices for local runtimes (#1675)

2026-06-16 10:14:49 +02:00

policies.mdx

fix(sandbox): restore GPU procfs baseline (#1522)

2026-06-03 11:08:21 +02:00

policy-advisor.mdx

feat(policy): add agentic approval loop (#1528)

2026-05-29 21:11:31 -07:00

providers-v2.mdx

feat(providers): support SPIFFE-backed token grants (#1784)

2026-06-10 10:54:39 -07:00