Ultra and Unified CCL

Apache Iggy: Hyper-Efficient Message Streaming at Laser Speed

Virtualized Elastic KV Cache for Dynamic GPU Sharing and Beyond

Zero-copy deserialization framework for Rust

KV cache store for distributed LLM inference

An embedded key-value database in pure Rust

AWS Libfabric

A library for mechanistic interpretability of GPT-style language models

The best ChatGPT that $100 can buy.

Optimized primitives for collective multi-GPU communication

A low level, low latency library, which can be used to accelerate network messages using shared memory and RDMA

LLM Attentions for Rust

Minimalist vLLM implementation in Rust

Heterogeneous AI Computing Virtualization Middleware(Project under CNCF)

Reference implementations of MLPerf® inference benchmarks

Rust Bootcamp Syllabus

Open MPI main development repository

Super-project for modularized Boost

Fast and memory-efficient exact attention

Performance-portable, length-agnostic SIMD with runtime dispatch

Multifunctional cross-platform interprocess communication toolkit for Rust.

A multiprocess drop-in replacement for Rust channels

A realtime-safe single-producer single-consumer (SPSC) ring buffer

A Rust client wrapper for Aeron C bindings

Rust aeron port

MPI Microbenchmarks

Triton backend that enables pre-process, post-processing and other logic to be implemented in Python.