llama.cpp · b11905

vulkan: rms norm optimization using subgroup reductions (#29882) * RMS Norm Subgroup reductions optimizaton * Fix build error * Improved optimization * Make usage of BLOCK_SIZE consistent * Add fast path to skip 2nd subgroupAdd * Add AMD Vendor to subgroup reduction opt * Add comment clarifying rms_dict and merge_maps behavior * Add unroll * Rebase * Fixed syntax issue * Add 128/256/512 block sizes to non-subgroup path * Remove trailing whitespace **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b11905/llama-b11905-bin-macos-arm64.tar.gz) - macOS Apple Silicon (arm64, KleidiAI enabled) [DISABLED](https://github.com/ggml-org/llama.cpp/pull/23780) - [macOS Intel (x64)](https://github.com/ggml-org/llama.c

1 个独立来源 · 1 篇报道切换最早在前 ↕

2026-10-27

llama.cpp releases03:08

b11905

vulkan：使用子组规约进行 RMS 法向量优化（#29882）* RMS 法向量子组规约优化 * 修复构建错误 * 改进优化效果 * 统一 BLOCK_SIZE 的使用方式 * 添加快速路径以跳过第 2 个子组 * 在子组规约优化中添加 AMD 厂商信息 * 添加注释，阐明 rms_dict 和 merge_maps 的行为 * 添加展开优化 * 重新变基 * 修复语法问题 * 为非子组路径新增 128/256/512 块大小 * 删除尾部空白 **网站：** - **证明文件：** - **macOS/iOS：** - [macOS Apple Silicon（arm64）](https://github.com/ggml-org/llama.cpp/releases/download/b11905/llama-b11905-bin-macos-arm64.tar.gz) - macOS Apple Silicon（arm64，启用 KleidiAI）[已禁用](https://github.com/ggml-org/llama.cpp/pull/23780) - [macOS Intel（x64）](https://github.com/ggml-org/llama.c)

事件进展与其他报道