ARM-software / kleidiaiLinks
This repository is a read-only mirror of https://gitlab.arm.com/kleidi/kleidiai
☆51Updated this week
Alternatives and similar repositories for kleidiai
Users that are interested in kleidiai are comparing it to the libraries listed below
Sorting:
- High-speed GEMV kernels, at most 2.7x speedup compared to pytorch baseline.☆109Updated 11 months ago
- Standalone Flash Attention v2 kernel without libtorch dependency☆110Updated 9 months ago
- Benchmark code for the "Online normalizer calculation for softmax" paper☆94Updated 6 years ago
- TileFusion is an experimental C++ macro kernel template library that elevates the abstraction level in CUDA C for tile processing.☆90Updated 2 weeks ago
- Several optimization methods of half-precision general matrix vector multiplication (HGEMV) using CUDA core.☆62Updated 9 months ago
- ☆75Updated 5 months ago
- llama INT4 cuda inference with AWQ☆54Updated 5 months ago
- An experimental CPU backend for Triton☆126Updated 3 weeks ago
- ☆60Updated 2 months ago
- ⚡️Write HGEMM from scratch using Tensor Cores with WMMA, MMA and CuTe API, Achieve Peak⚡️ Performance.☆80Updated last month
- Fast Hadamard transform in CUDA, with a PyTorch interface☆201Updated last year
- ⚡️FFPA: Extend FlashAttention-2 with Split-D, achieve ~O(1) SRAM complexity for large headdim, 1.8x~3x↑ vs SDPA.☆186Updated last month
- We invite you to visit and follow our new repository at https://github.com/microsoft/TileFusion. TiledCUDA is a highly efficient kernel …☆183Updated 4 months ago
- PyTorch extension for emulating FP8 data formats on standard FP32 Xeon/GPU hardware.☆110Updated 6 months ago
- ☆149Updated 2 years ago
- DeeperGEMM: crazy optimized version☆69Updated last month
- An experimental CPU backend for Triton (https//github.com/openai/triton)☆43Updated 3 months ago
- QQQ is an innovative and hardware-optimized W4A8 quantization solution for LLMs.☆128Updated 2 months ago
- ☆71Updated last month
- ☆29Updated 4 months ago
- ☆38Updated this week
- This is a demo how to write a high performance convolution run on apple silicon☆54Updated 3 years ago
- CUDA Matrix Multiplication Optimization☆196Updated 11 months ago
- ☆96Updated 9 months ago
- ☆81Updated 7 months ago
- A standalone GEMM kernel for fp16 activation and quantized weight, extracted from FasterTransformer☆92Updated 3 weeks ago
- ☆34Updated last year
- Ahead of Time (AOT) Triton Math Library☆66Updated last week
- ☆69Updated 7 months ago
- ☆86Updated 3 months ago