Fast and memory-efficient exact attention

Performance of the C++ interface of flash attention and flash attention v2 in large language model (LLM) inference scenarios.

ROCm BLAS marshalling library

how to optimize some algorithm in cuda.

flash attention tutorial written in python, triton, cuda, cutlass

FlashAttention2 implementation with TensorCore WMMA API

Implement FlashAttention v2 with minimal code to learn.

A flash attention2 extension for stable diffusion webui in Linux pytorch-rocm environments.

使用 cutlass 实现 flash-attention 精简版，具有教学意义

An unofficial cuda assembler, for all generations of SASS, hopefully ：）

Flash Attention in raw Cuda C beating PyTorch

A framework that support executing unmodified CUDA source code on non-NVIDIA devices.

Codes & examples for "CUDA - From Correctness to Performance"

LLVM/MLIR based compiler instrumentation of AMD GPU kernels

CUDA on non-NVIDIA GPUs

AMD ROCm™ Software - GitHub Home

CuPBoP-AMD is a CUDA translator that translates CUDA programs at NVVM IR level to HIP-compatible IR that can run on AMD GPUs. Currently, CuPBoP-AMD translates a broader range of applications in the…

HIP: C++ Heterogeneous-Compute Interface for Portability

Implementation of a simple CNN using CUDA

CuPBoP-AMD is a CUDA translator that translates CUDA programs at NVVM IR level to HIP-compatible IR that can run on AMD GPUs.

A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper and Ada GPUs, to provide better performance with lower memory utilizatio…

A collection of pre-trained, state-of-the-art models in the ONNX format

AutoKernel 是一个简单易用，低门槛的自动算子优化工具，提高深度学习算法部署效率。

Eigen is a C++ template library for linear algebra: matrices, vectors, numerical solvers, and related algorithms.

Protocol Buffers - Google's data interchange format