DeepGEMM-Ascend: Fast GEMM Kernels for Huawei Ascend NPUs
DeepGEMM-Ascend ports DeepSeek's DeepGEMM library to Huawei Ascend NPUs, with API-compatible kernels for BF16, FP8 and FP4 GEMM, MQA logits, and MegaMoE. It targets teams optimizing inference and training workloads on Ascend 950 hardware.
NewHealth 44/100
C++LibraryPerformance
Added Nov 14, 2026 View details
