DeepSeek open-sources Ascend ports of DeepGEMM and DeepEP, with Huawei backing
DeepSeek has open-sourced two new repositories for Huawei's Ascend accelerators: DeepGEMM-Ascend for matrix multiplication kernels and DeepEP-Ascend for the all-to-all communication that mixture-of-experts models rely on, with Huawei's support. TechRadar notes that the other components named in the announcement — TileKernels, DeepSelect and FlashMLA — are Ascend-targeted updates to existing projects, and that the headline TileLang is neither new nor DeepSeek's own. Requirements are narrow: DeepGEMM-Ascend needs Ascend 950 series silicon and CANN 9.20, DeepEP-Ascend needs Ascend 950 with UBMEM connectivity, and the published benchmarks ran on a proof-of-concept hardware kit DeepSeek says is not a publicly distributed release.
It is an attempt by Chinese vendors to fill in the low-level training and inference kernels outside Nvidia's CUDA ecosystem, though Ascend 950 availability keeps the near-term impact mostly domestic.