Skip to content
AI Primer

DeepGEMM

clean and efficient BLAS kernel library on GPU

DeepGEMM is DeepSeek's open-source, high-performance CUDA tensor-core/BLAS kernel library for modern LLM computation primitives including GEMMs, fused MoE, MQA scoring, HyperConnection, and related GPU kernels compiled through a lightweight JIT module.

Screenshot of DeepGEMM website

Recent stories

0 linked stories
No linked stories yet.
AI PrimerAI Primer

Your daily guide to AI tools, workflows, and creative inspiration.

© 2026 AI Primer. All rights reserved.