Search

Optimized primitives for collective multi-GPU communication

Multi-GPU CUDA stress test

Examples demonstrating available options to program multiple GPUs in a single node or a cluster

Large-scale, multi-GPU capable, kernel solver

AMD RAD's multi-GPU Triton-based framework for seamless multi-GPU programming

Kubernetes operator for self-hosted LLM inference across a heterogeneous GPU fleet: NVIDIA CUDA, AMD Vulkan, and Apple Silicon Metal. Runtimes: llama.cpp, vLLM, TGI, mlx-server. Multi-GPU sharding, model caching, OpenAI-compatible endpoints. Apache-2.0, run across homelab and on-prem fleets, actively developed.

mKernel: fast multi-node, multi-GPU fused kernels

SSD-streaming inference engine for giant MoE models (Rust + CUDA). GLM 5.2 743B at 2 tok/s and Hy3 295B at 7 tok/s on two consumer 16GB GPUs. Zero-config multi-GPU: measures PCIe bandwidth, places attention and hot experts where they fit.