PulseAugur
中
实时 18:22:32

llama.cpp b11786 版本改进 CPU 内核性能

llama.cpp 项目发布了 b11786 版本,更新了其 ggml-cpu 计算内核。这些更改侧重于通过移除对 fp16 表的需求来改进 f32 gelu_quick 和 geglu_quick 函数的计算,并引入了 VXE gelu_quick 路径。该版本还扩展了这些 CPU 内核的测试,并指出其兼容 macOS、Linux、Android 和 Windows 等多种操作系统,特别提到了 Apple Silicon 和 openEuler。 AI

影响 llama.cpp 计算内核的性能提升可能导致兼容模型的本地推理速度加快。

排序理由 特定项目(llama.cpp)的软件发布,改进了计算内核的性能。

在 llama.cpp — Releases 阅读 →

AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →

llama.cpp b11786 版本改进 CPU 内核性能

本文如何被排名

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
特定项目(llama.cpp)的软件发布,改进了计算内核的性能。
Source corroboration
3 independent sources
Multiple independent publishers reporting the same story raises confidence that it's real and newsworthy.
Topics
infra
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
2 days old
Coverage has settled into its steady-state source set.
Coverage growth since scoring
+1 source(s) since last score
New sources have picked up this story since our last re-score. Score will update on the next scoring pass.

完整方法见我们的编辑标准。

报道来源 [3]

  1. Ollama — Releases TIER_1 (SK) · github-actions[bot] ·

    v0.40.10

    <h2>NVIDIA Nemotron 3.5 Bolt E4B</h2> <p><a href="https://ollama.com/library/nemotron-3.5-bolt" rel="nofollow">Nemotron 3.5 Bolt E4B</a> is NVIDIA’s smaller, Gemma 4-based vision-language model for fast local conversations, image understanding, and tool use. It supports a 128K-to…

  2. llama.cpp — Releases TIER_1 (SO) · github-actions[bot] ·

    b11786

    <details open=""> <p>ggml-cpu : compute f32 gelu_quick and geglu_quick without the fp16 table (<a class="issue-link js-issue-link" href="https://github.com/ggml-org/llama.cpp/pull/30737">#30737</a>)</p> <ul> <li> <p>ggml-cpu : compute f32 gelu_quick and geglu_quick without the fp…

  3. Medium — Claude tag TIER_1 Norsk(NO) · rhythm ·

    Claude Sonnet 5.5: Cheaper Per Task, Not Cheaper Per Token

    <div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@susanwatson1931/claude-sonnet-5-5-cheaper-per-task-not-cheaper-per-token-513e0fa11fa9?source=rss------claude-5"><img src="https://cdn-images-1.medium.com/max/1536/0*KqLMUuUCjrCtQtc3.png" width…