deepseek-ai / DeepEP
DeepEP: an efficient expert-parallel communication library
{{ message }}
See what the GitHub community is most excited about today.
DeepEP: an efficient expert-parallel communication library
NCCL Tests
DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling
GPU accelerated decision optimization
Tile primitives for speedy kernels
RTP-LLM: Alibaba's high-performance LLM inference engine for diverse applications.
Causal depthwise conv1d in CUDA, with a PyTorch interface
CUDA Kernel Benchmarking Library
Cuda mesh utils.
RAPIDS Accelerator JNI For Apache Spark
Instant neural graphics primitives: lightning fast NeRF and more
[ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x compared to FlashAttention, without losing end-to-end metrics across language, image, and video models.
[ARCHIVED] Cooperative primitives for CUDA C++. See https://github.com/NVIDIA/cccl