v20261001.1 — vLLM 0.29.0+rv.mxfp8.mtp.0a6739845a32.965e748f74fe.c6f1c3f3e582

Linux arm64 (SM12x family) and amd64 (sm_120) image index, vLLM 0.29.0+rv.mxfp8.mtp.0a6739845a32.965e748f74fe.c6f1c3f3e582.

docker pull ghcr.io/randomvariable/vllm-b12x-multi@sha256:43aaf6d1b5e0b96bff630f5b5211afccd1b4838c40a29864aaee3eb7dae6cbde

Field	Value
Digest	sha256:43aaf6d1b5e0b96bff630f5b5211afccd1b4838c40a29864aaee3eb7dae6cbde
Release tag	ghcr.io/randomvariable/vllm-b12x-multi:v20261001.1
Publication tag	ghcr.io/randomvariable/vllm-b12x-multi:vllmb12x-dev-rv-mxfp8-mtp-0a6739845a32-607a9619a9c3-20261001-n70
vLLM	0a6739845a3249b07a30ad9b1e720e5e3fb6236f
B12X	965e748f74fe9dac0a35c497a0947c5b108617c5
Builder	607a9619a9c31e7b75d776e0ee5d398b8067fe2e

Base vLLM image additions

• Mooncake Transfer Engine CUDA 13 0.3.13.post1 provides the mooncake Python package and its service and benchmark commands for deployments that select vLLM's MooncakeStoreConnector. It does not enable a connector or start Mooncake services.
• See Mooncake Transfer Engine for the installed surface and deployment boundary.

Included upstream changes

The image is built from pinned fork revisions rather than from upstream branches, so this table records which upstream changes the current lock carries. Regenerate it with scripts/vllmb12x-included-changes.py.

Component	Change	Included as
vLLM 0a6739845a32	Native MXFP8 MTP through the ModelOpt B12X backend (no upstream pull request)	Merged into the pinned revision (4 commits)
vLLM 0a6739845a32	Partial port of vLLM PR 779: Qwen HC ownership and checkpoint coalescing (no upstream pull request)	Merged into the pinned revision (9 commits)
vLLM 0a6739845a32	local-inference-lab/vllm#800 Bounded shared-memory broadcast waits	Merged into the pinned revision (1 commits)
vLLM 0a6739845a32	Rewrite flash_attn.cute imports to vllm.vllm_flash_attn.cute (no upstream pull request)	Applied at build time by third_party/vllm_flash_attn_cute_namespace.patch
B12X 965e748f74fe	Native block-scaled MXFP8 W8A8 MoE execution (no upstream pull request)	Merged into the pinned revision (3 commits)
B12X 965e748f74fe	PLE checkpoint exports, prepared launcher closures and collective entry barriers (no upstream pull request)	Merged into the pinned revision (11 commits)

Serving verification (2 GB10 groups x TP=2, idle, image above)

llm-inference-bench 0.7.5 against one group's rank-zero endpoint, 30 s cells, one sample per cell, temperature 1, max 8,192 output tokens. Qwen3.8-Flash-Next-NVFP4 qad-step5500-ple1000, MTP depth 3, --max-num-seqs 8, utilization 0.74.

• Backend selection: B12X (NVFP4 target experts), B12X_MXFP8 (MXFP8 draft experts); nothing on Marlin. 72/72 B12X preparation units ready.
• Prefill: 2,857 / 2,882 / 2,664 / 2,416 tok/s at 8K / 16K / 64K / 128K (TTFT 2.87 / 5.61 / 24.07 / 52.98 s). The 128K cell ranged 1,873-2,416 tok/s across three single samples.
• Decode, aggregate tok/s / engine steps/s (accept length): C1 52.5 / 23.4 (2.24); C4 126.7 / 59.3 (2.14); C8 181.8 / 86.8 (2.09). Context 0 and 16K agree within ~3% on steps/s.
• Not measured: gateway-path overhead, prefix-cache reuse, comparison against Marlin.

v20260917.1 — vLLM 0.1.dev1+vllmb12x.gbd22e0f25043

Linux ARM64 image for NVIDIA DGX Spark (GB10, sm_121a), vLLM 0.1.dev1+vllmb12x.gbd22e0f25043.

docker pull ghcr.io/randomvariable/vllm-b12x-multi@sha256:cdf2602bb2e42da5b8f8cae8ae4bbe0842388cd6d8bf6ca2e6e3c27d7e27ad1e

Field	Value
Digest	sha256:cdf2602bb2e42da5b8f8cae8ae4bbe0842388cd6d8bf6ca2e6e3c27d7e27ad1e
Release tag	ghcr.io/randomvariable/vllm-b12x-multi:v20260917.1
Publication tag	ghcr.io/randomvariable/vllm-b12x-multi:vllmb12x-dev-rv-jovian-judgement-profile-base-bd22e0f25043-13527e984717-20260917-n59
vLLM	bd22e0f250439659d9fe30903b7aa124c321a932
B12X	4401ce4fb98265c3b1e1579ae03960ff41d84d50
Builder	13527e9847178f123098b7967c27333f4c82773b

Included upstream changes

The image is built from pinned fork revisions rather than from upstream branches, so this table records which upstream changes the current lock carries. Regenerate it with scripts/vllmb12x-included-changes.py.

Component	Change	Included as
vLLM bd22e0f25043	local-inference-lab/vllm#777 fix(qwen): propagate MTP positional overrides	Merged into the pinned revision (3 commits)
vLLM bd22e0f25043	local-inference-lab/vllm#779 perf(qwen): shard TP4 HC prefill and coalesce recurrent checkpoints	Merged into the pinned revision (9 commits)
vLLM bd22e0f25043	vllm-project/vllm#52917 Adaptive spin grace and bounded architectural waits for shm_broadcast	Applied at build time by third_party/vllm_shm_broadcast_spin_grace.patch
vLLM bd22e0f25043	Rewrite flash_attn.cute imports to vllm.vllm_flash_attn.cute (no upstream pull request)	Applied at build time by third_party/vllm_flash_attn_cute_namespace.patch
B12X 4401ce4fb982	local-inference-lab/b12x#384 fix(preparation): retain prepared launchers and coordinate collectives	Merged into the pinned revision (10 commits)
B12X 4401ce4fb982	local-inference-lab/b12x#386 feat(ple): export prepared internal prefill checkpoints	Merged into the pinned revision (2 commits)
B12X 4401ce4fb982	local-inference-lab/b12x#387 perf(qsa): reuse representative keys across paired queries	Merged into the pinned revision (4 commits)