Repository navigation
blog: speculative decoding, smaller files, one GPU many users - #43
Merged
Merged
Conversation
…h, ROCmFPX, a ternary 27B) Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
bong-water-water-bong
changed the title
bong-water-water-bong
enabled auto-merge (squash)
September 24, 2026 18:51
bong-water-water-bong
added a commit
that referenced
this pull request
Sep 29, 2026
…B 97% of Vulkan, 74 C) (#236) llama.cpp #43: Q3_K and IQ3_S in the K-quant decode kernels (Qwen3.8-27B UD-Q4_K_XL 11.60 -> 12.15 tok/s, Vulkan 12.48). llama.cpp #44: long stream waits sleep instead of spinning a CPU core in ROCr's signal wait (27B decode 71-76 C / ~125 W instead of 83-93 C / 133-137 W, -0.4%). docs/hrx.md updated. Co-authored-by: bong-water-water-bong <bong-water-water-bong@1bit.gg> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
This was referenced Oct 7, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Two posts from today's measurements on Strix Halo; every number is from the runs behind #37, #41 and #42.
--mtp-max 6 --mtp-p-min 0.5. Qwen3-Coder-30B-A3B (3B active): no draft model beats no drafting across 36 combinations.Please read them before merging: they are written in the project's voice. A third post, on serving many users at once (batching and
--adaptive), follows when its last load test is in.🤖 Generated with Claude Code