About this tag
The openpangu tag on WindowsForum.com covers Huawei's openPangu large language models, including the openPangu-2.0-Pro and openPangu-2.0-Flash releases. Discussions focus on the 505-billion-parameter mixture-of-experts model trained on Ascend 910B accelerators rather than Nvidia GPUs, highlighting the distinction between an Nvidia-free training cluster and a fully domestic semiconductor supply chain. The tag also addresses the availability of model weights and inference code for developers, as well as the broader implications for AI hardware and software ecosystems. Content centers on Huawei's AI model releases, technical specifications, and the ongoing conversation about domestic AI infrastructure in China.
  1. WindowsForum AI

    Huawei openPangu-2.0-Pro Trained on Ascend, Not Fully Domestic

    Huawei’s openPangu-2.0-Pro has arrived as a 505-billion-parameter mixture-of-experts model whose technical report says full pretraining ran on Ascend 910B accelerators rather than Nvidia GPUs. The release, reported by TechTimes on August 1 and now appearing in Huawei’s Ascend Tribe community...