Back to News

DeepSeek to deploy 160K+ Huawei Ascend 950DT chips at Inner Mongolia data center

#deepseek#huawei#ai-chips#datacenter

Bloomberg sources report that DeepSeek plans to deploy at least 160,000 Huawei Ascend 950DT accelerators at a data center it is building in Inner Mongolia, which would create one of the largest Huawei chip clusters. The deployment underscores DeepSeek's reliance on Huawei hardware amid ongoing US export restrictions.

Coverage timeline

  1. Techmeme
    Sources: DeepSeek plans to use 160K+ of Huawei's Ascend 950DT chips at an Inner Mongolia data center, which would create one of the largest Huawei chip clusters (Mackenzie Hawkins/Bloomberg)

    Mackenzie Hawkins / Bloomberg : Sources: DeepSeek plans to use 160K+ of Huawei's Ascend 950DT chips at an Inner Mongolia data center, which would create one of the largest Huawei chip clusters — DeepSeek plans to deploy at least 160,000 of Huawei Technologies Co.'s top accelerators at a massive data center it's building …

Related stories

  • Tencent signs ~$7B Oracle deal for 100K AI chips via Southeast Asia

    Sources report Tencent signed a five-year deal worth about $7 billion with Oracle this year, granting access to roughly 100,000 advanced AI chips unavailable in China, delivered via Oracle's Southeast Asian data centers. The move comes amid an intensifying AI race with ByteDance and Alibaba.

    1 reports · 1 sources

  • DeepSeek open-sources Ascend infrastructure components, partners with Huawei

    DeepSeek announced on September 30 the open-sourcing of infrastructure components for Huawei's Ascend platform, including the TileLang compiler, high-performance computing libraries (DeepGEMM, FlashMLA, TileKernels, DeepSelect), and the DeepEP distributed communication library. Huawei also open-sourced joint innovations around Ascend 950 and supernodes in the CANN community, covering deployment, training, KV cache pooling, and Agentic RL. The move aims to build an efficient AI chip software ecosystem with Huawei, providing developers with tools corresponding to those previously released for GPU platforms.

    3 reports · 3 sources

  • Inferact runs Kimi K3 on 16 Google TPU v7 at 709 tokens/s, 57% faster than GB200

    Inferact, a startup founded by the original vLLM team, reported that 16 Google TPU v7 chips ran Kimi K3 at 709 tokens per second, 57% faster than 16 Nvidia GB200 GPUs under identical conditions. The speedup comes from a custom megakernel inference kernel and DeepSeek's DSpark speculative decoding framework; the code has been open-sourced.

    1 reports · 1 sources

  • DeepSeek CEO: Training on Huawei chips is a major bet; Huawei to deliver training chips by Q1 2027

    The Information reports that DeepSeek CEO Liang Wenfeng told investors that using more domestic chips to train models is a major priority, calling training on Huawei chips one of the company's biggest bets. Huawei is expected to deliver training chips in Q4 2026 or Q1 2027, according to sources.

    1 reports · 1 sources

News updates every hour. Subscribe in your RSS reader to get new stories automatically.

Subscribe