本文へ移動
AI NEWS JOURNAL
llama.cpp releasesAI NEWS / JOURNAL

b12621

A patch optimizing SYCL/drama 3 coordination for CONV_TRANSPOSE_1D reduces execution time significantly across various input and kernel sizes, with speedups up to 710.23x on Arc Pro B70 using oneAPI 2026.1. The optimization maintains correctness with identical output results, though test failures persist due to a non-F32 SWIGLU_OAI assertion unrelated to the change. Performance gains are validated in a chained 4-layer stack scenario, reducing processing time from hundreds of milliseconds to under 30ms.

翻訳サービスが停止中のため原文を表示しています。

<details open> sycl: bound CONV_TRANSPOSE_1D input iteration (#31109) Used angelX to optimize SYCL/drama 3 coordination, results are as follows Each output reads only inputs with i*s0 <= idx < i*s0 + K, in the same ascending order as before. Arc Pro B70, oneAPI 2026.1, test-backend-ops perf, base 9b56ccf00 vs patch, median of 3 same-boot A/B/A rounds, A/A 1.000-1.001: | input [ne0, ne1] | kernel [K, OC, IC] | s0 | base us | patch us | speedup | |---|---|---|---|---|---| | [16, 1536] | [16, 768, 1536] | 8 | 3910.57 | 1348.59 | 2.90x | | [128, 768] | [16, 384, 768] | 8 | 37735.00 | 1665.26 | 22.66x | | [1024, 384] | [8, 192, 384] | 4 | 270371.75 | 1471.74 | 183.71x | | [4096, 192] | [4, 96, 192] | 2 | 534614.63 | 752.74 | 710.23x | Shapes are the drama 3 codec decoder upsampling stack, added as perf cases. Chained 4-layer stack at 1/4/24 input frames: base 66.3/847.4/28638.5 ms, patch 2.3/6.7/30.7 ms. Kernel numbers, not end-to-end TTS latency. Correctness: test-backend-ops -o CONV_TRANSPOSE_1D 116/116 on base and patch, and 116/116 rebased on 3f03b1772. Seeded output dump (7.46 MB) byte-identical to base. Full test-backend-ops aborts on base and patch alike at the non-F32 SWIGLU_OAI assert (element_wise.cpp:1187, see #30857). With only those cases excluded: 20145/20173 on base and patch, and every patch-run failure also fails on base. Co-authored-by: Titaniumtown <titaniumtown@proton.me> </details> **Website:** - <https://llama.app> **Attestations:** - <https://github.com/ggml-org/llama.cpp/attestations/70644894> **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-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.cpp/releases/download/b12621/llama-b12621-bin-macos-x64.tar.gz) - [iOS XCFramework](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-xcframework.zip) **Linux:** - [Ubuntu x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-ubuntu-x64.tar.gz) - [Ubuntu arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-ubuntu-arm64.tar.gz) - [Ubuntu 24.04 s390x (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-ubuntu-s390x.tar.gz) - [Ubuntu 26.04 s390x (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-ubuntu-26.04-s390x.tar.gz) - [Ubuntu 26.04 s390x (zDNN)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-ubuntu-zdnn-s390x.tar.gz) - [Ubuntu x64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-ubuntu-vulkan-x64.tar.gz) - [Ubuntu arm64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-ubuntu-vulkan-arm64.tar.gz) - [Ubuntu x64 (CUDA 12)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-ubuntu-cuda-12.8-x64.tar.gz) - [CUDA 12.8 libraries](https://github.com/ggml-org/llama.cpp/releases/download/b12621/cudart-llama-b12621-bin-ubuntu-cuda-12.8-x64.tar.gz) - [Ubuntu x64 (CUDA 13)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-ubuntu-cuda-13.5-x64.tar.gz) - [CUDA 13.5 libraries](https://github.com/ggml-org/llama.cpp/releases/download/b12621/cudart-llama-b12621-bin-ubuntu-cuda-13.5-x64.tar.gz) - [Ubuntu arm64 (CUDA 13)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-ubuntu-cuda-13.5-arm64.tar.gz) - [CUDA 13.5 libraries](https://github.com/ggml-org/llama.cpp/releases/download/b12621/cudart-llama-b12621-bin-ubuntu-cuda-13.5-arm64.tar.gz) - [Ubuntu x64 (ROCm 10.0)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-ubuntu-rocm-10.0-x64.tar.gz) - [Ubuntu x64 (OpenVINO)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-ubuntu-openvino-2026.4.1-x64.tar.gz) - [Ubuntu x64 (SYCL FP32)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-ubuntu-sycl-fp32-x64.tar.gz) - [Ubuntu x64 (SYCL FP16)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-ubuntu-sycl-fp16-x64.tar.gz) - [Linux arm64 (Snapdragon: CPU, Adreno GPU, Hexagon NPU)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-linux-arm64-snapdragon.tar.gz) - [setup guide](https://github.com/ggml-org/llama.cpp/blob/master/docs/backend/snapdragon/linux.md) **Android:** - [Android arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-android-arm64.tar.gz) - [Android arm64 (Snapdragon: CPU, Adreno GPU, Hexagon NPU)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-android-arm64-snapdragon.tar.gz) - [setup guide](https://github.com/ggml-org/llama.cpp/blob/master/docs/backend/snapdragon/README.md) **Windows:** - [Windows x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-win-cpu-x64.zip) - [Windows arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-win-cpu-arm64.zip) - [Windows arm64 (OpenCL Adreno)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-win-opencl-adreno-arm64.zip) - [Windows x64 (CUDA 12)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-win-cuda-12.4-x64.zip) - [CUDA 12.4 DLLs](https://github.com/ggml-org/llama.cpp/releases/download/b12621/cudart-llama-bin-win-cuda-12.4-x64.zip) - [Windows x64 (CUDA 13)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-win-cuda-13.5-x64.zip) - [CUDA 13.5 DLLs](https://github.com/ggml-org/llama.cpp/releases/download/b12621/cudart-llama-bin-win-cuda-13.5-x64.zip) - [Windows arm64 (CUDA 13)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-win-cuda-13.5-arm64.zip) - [CUDA 13.5 DLLs](https://github.com/ggml-org/llama.cpp/releases/download/b12621/cudart-llama-bin-win-cuda-13.5-arm64.zip) - [Windows x64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-win-vulkan-x64.zip) - [Windows arm64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-win-vulkan-arm64.zip) - [Windows x64 (OpenVINO)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-win-openvino-2026.4.1-x64.zip) - [Windows x64 (SYCL)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-win-sycl-x64.zip) - [Windows x64 (ROCm 10.0)](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-bin-win-rocm-10.0-x64.zip) **openEuler:** - [DISABLED](https://github.com/ggml-org/llama.cpp/pull/23705) - openEuler x86 (310p) - openEuler x86 (910b, ACL Graph) - openEuler aarch64 (310p) - openEuler aarch64 (910b, ACL Graph) **UI:** - [UI](https://github.com/ggml-org/llama.cpp/releases/download/b12621/llama-b12621-ui.tar.gz)

出典

原文を読む ↗

関連テーマ

← すべてのニュースへ