[II] Enable B12X DCP collectives at world size 16

[voipmonitor]

Status

Qualified.

Behavior

B12X query-head gather, LSE reduction, and graph warmup accept DCP world sizes 2, 4, 8, and 16. Other world sizes continue to use the existing NCCL fallback.

Technical reason

B12X DcpAllToAllPool supports world size 16, but the vLLM dispatch guards excluded it. The guard prevented TP16/DCP16 serving from selecting the B12X DCP transport even when its tensor geometry satisfied the runtime contract.

Compatibility

No model-specific geometry, numerical operation, tensor format, channel ownership, or fallback behavior changes. TP2, TP4, and TP8 retain the same path.

Validation

• Ruff formatting and checks pass.
• git diff --check passes.
• 37 relevant tests from tests/distributed/test_dcp_a2a.py pass in the CUDA 13.3/PyTorch 2.13 Kimi-K3 image.
• Explicit coverage exercises DCP16 query gather, LSE reduction, and warmup; world size 6 still verifies the NCCL fallback.

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

🗂️ Base branches to auto review (1)

• dev/*

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 860c9c39-032b-49ac-b956-81508b2c0aa7

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[voipmonitor]

The behavior implemented by this pull request is preserved unchanged in vLLM #384 as commit 31aae805f3d3. Stable Git patch IDs match. Review and merge #384; this pull request is closed to avoid duplicate review.