aiBACKGROUND
b11730: cuda: cap MMQ Stream-K blocks at available work (#30553)
BACKGROUND
1 SIGNALFIRST DETECTED 19 October 2026
Narrative synthesis pending — 1 signal collected.
How the story developed
19 Oct
b11730: cuda: cap MMQ Stream-K blocks at available work (#30553)
REPORT · WHAT · GitHub: llama.cpp
1.8
Other feeds and industry voices
17 Oct
b11713: CUDA: faster MMVQ for low-bit types (#30517)
<ul> <li> <p>CUDA: faster MMVQ for low-bit types</p> </li> <li> <p>ggml-cuda: hip: small nwarps GCN fix on Q3_K(ncols=2,3), Q6_K(ncols=4)</p> </li> </ul> <hr /> <p>Co-authored-by:
REPORT · WHAT · GitHub: llama.cppOPEN STORY
15 Oct
b11632: cuda, hip: GDN chunked kernel (#29353)
<ul> <li>cuda, hip: GDN chunked kernel</li> </ul> <p>Ported-from: <a class="issue-link js-issue-link"
REPORT · WHAT · GitHub: llama.cppOPEN STORY
Sources
GitHub: llama.cpp
NewsHive monitors these sources continuously. All signal titles above link to the original reporting.
Intelligence by NewsHive. Need help navigating what this means for your business? Contact GeekyBee →