Amd Python Packages

vllm

A high-throughput and memory-efficient inference and serving engine for LLMs

pyopencl

OpenCL integration for Python, plus shiny features

dstack

Vendor-agnostic orchestration for training, inference and agentic workloads across NVIDIA, AMD, TPU, and Tenstorrent on clouds, Kubernetes, and bare metal.

lmcache

LMCache: Supercharge Your LLM with the Fastest KV Cache Layer

quadrants

High-performance multi-platform compiler for physics simulation

vllm-tpu

A high-throughput and memory-efficient inference and serving engine for LLMs

conch-triton-kernels

A "standard library" of Triton kernels.

vllm-cpu-nightly

A high-throughput and memory-efficient inference and serving engine for LLMs

gstaichi

High-performance multi-platform compiler for physics simulation

amd-gaia

Build AI agents for your PC

zenmaster

Adjust power management settings for Ryzen CPUs on Linux, Windows, and macOS

lmcache-cli

LMCache: Supercharge Your LLM with the Fastest KV Cache Layer

roctop

`roctop` is a small AMD ROCm GPU monitor. It shows GPU VRAM usage, utilization, temperature, and GPU processes in a refreshing terminal UI.

eyepy

A python package to read, analyse and visualize OCT and fundus data from various sources.

uccl

UCCL: Ultra and Unified CCL

amdfan

Fan monitor and controller for AMD gpus in Linux

nabla-ml

Nabla: High-Performance Scientific Computing

amd-smi-wsl

Drop-in amdsmi for AMD GPUs on WSL2 (DirectX/dxgkrnl passthrough) where the native amdsmi is unavailable — unblocks vLLM, PyTorch ROCm, GPU monitoring/tooling, etc.

sovereign-inference

Ultra-fast LLM inference engine with a Vulkan compute backend

grilly

GPU-accelerated neural network operations using Vulkan compute shaders.

eyepie

A python package to read, analyse and visualize OCT and fundus data from various sources.

onnx-web

web UI for GPU-accelerated ONNX pipelines like Stable Diffusion, even on Windows and AMD

gputop

A simple real-time GPU monitoring tool for NVIDIA, AMD and Intel GPUs.

ai-dynamo-vllm

A high-throughput and memory-efficient inference and serving engine for LLMs