High-performance CUDA kernels for Kimi Delta Attention built on CUTLASS
High-performance GPU kernel library for blazing fast LLM inference and serving
Ultra-high-performance, secure, all-in-one acceleration engine for developer resources
A flexible framework for cutting-edge LLM inference and fine-tuning optimizations with CPU-GPU heterogeneous computing
Postgres rewritten in Rust with AI assistance, passing 100% of regression tests and delivering 50% faster performance
Agent skills for coding agents to build modern, performant web apps
