DeepGEMM is a clean, unified, high-performance CUDA library specifically engineered to serve the core matrix computation primitives required by state-of-the-art large language models (such as DeepSeek ...
Welcome to the AI Applications in Accelerators Tutorial & Mini-Hackathon repository. This project hosts the computational challenge, simulation environment, benchmark notebooks, and documentation for ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results