ROCm / AITemplate
AITemplate is a Python framework which renders neural network into high performance CUDA/HIP C++ code. Specialized for FP16 TensorCore (NVIDIA GPU) and MatrixCore (AMD GPU) inference.
☆11Updated 9 months ago
Alternatives and similar repositories for AITemplate:
Users that are interested in AITemplate are comparing it to the libraries listed below
- Development repository for the Triton language and compiler☆114Updated this week
- Fast and memory-efficient exact attention☆163Updated this week
- OpenAI Triton backend for Intel® GPUs☆172Updated this week
- An efficient GPU support for LLM inference with x-bit quantization (e.g. FP6,FP5).☆243Updated 5 months ago
- Composable Kernel: Performance Portable Programming Model for Machine Learning Tensor Operators☆373Updated this week
- ☆26Updated this week
- ☆21Updated last month
- Ahead of Time (AOT) Triton Math Library☆56Updated 2 weeks ago
- hipBLASLt is a library that provides general matrix-matrix operations with a flexible API and extends functionalities beyond a traditiona…☆84Updated this week
- AMD's graph optimization engine.☆213Updated this week
- A high-throughput and memory-efficient inference and serving engine for LLMs☆70Updated this week
- a simple Flash Attention v2 implementation with ROCM (RDNA3 GPU, roc wmma), mainly used for stable diffusion(ComfyUI) in Windows ZLUDA en…☆37Updated 7 months ago
- ☆193Updated 8 months ago
- AI Tensor Engine for ROCm☆142Updated this week
- Intel® Extension for DeepSpeed* is an extension to DeepSpeed that brings feature support with SYCL kernels on Intel GPU(XPU) device. Note…☆62Updated 3 weeks ago
- RCCL Performance Benchmark Tests☆60Updated 3 weeks ago
- oneCCL Bindings for Pytorch*☆91Updated this week
- ROCm BLAS marshalling library☆136Updated this week
- ☆20Updated last week
- An experimental CPU backend for Triton (https//github.com/openai/triton)☆40Updated 2 weeks ago
- collection of benchmarks to measure basic GPU capabilities☆340Updated last month
- python package of rocm-smi-lib☆20Updated 6 months ago
- Test suite for probing the numerical behavior of NVIDIA tensor cores☆37Updated 8 months ago
- A tool for generating information about the matrix multiplication instructions in AMD Radeon™ and AMD Instinct™ accelerators☆80Updated this week
- High-speed GEMV kernels, at most 2.7x speedup compared to pytorch baseline.☆103Updated 8 months ago
- ☆90Updated 3 weeks ago
- ☆192Updated last week
- ROCm Communication Collectives Library (RCCL)☆314Updated this week
- Documentation for vLLM Dev Channel releases☆9Updated 3 months ago
- Shared Middle-Layer for Triton Compilation☆236Updated last week