PulseAugur
实时 22:32:09
English(EN) Huawei is attacking Nvidia’s software moat by open-sourcing its AscendNPU IR, the core of the BiSheng compiler for the upcoming Ascend 950 chip. By adding nativ

华为开源AscendNPU IR以挑战Nvidia的AI硬件主导地位

华为通过开源其BiSheng编译器核心组件AscendNPU IR,来挑战Nvidia在AI硬件市场的统治地位。此举旨在支持即将推出的Ascend 950芯片,并包含对Triton的原生支持,使开发人员无需依赖专有接口即可创建高性能内核。此举被视为一项战略举措,旨在促进中国AI开发实验室从Nvidia当前硬件架构的迁移。 AI

影响 华为开源其AscendNPU IR可能会促进一个更具竞争力的AI硬件生态系统,可能减少对单一供应商的依赖,并鼓励更广泛地采用其自身的芯片架构。

排序理由 公司开源核心技术以与主导者竞争。[lever_c_demoted from significant: ic=1 ai=0.7]

在 Mastodon — mastodon.social 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

华为开源AscendNPU IR以挑战Nvidia的AI硬件主导地位

报道来源 [1]

  1. Mastodon — mastodon.social TIER_1 English(EN) · asiaai ·

    Huawei is attacking Nvidia’s software moat by open-sourcing its AscendNPU IR, the core of the BiSheng compiler for the upcoming Ascend 950 chip. By adding nativ

    Huawei is attacking Nvidia’s software moat by open-sourcing its AscendNPU IR, the core of the BiSheng compiler for the upcoming Ascend 950 chip. By adding native Triton support, Huawei lets developers write high-performance kernels without learning proprietary APIs. It’s a crucia…