06/05/2025, 02:36 AM UTC
MLPerf训练基准测试v5.0结果发布MLPerf Training v5.0 is Out
➀ MLPerf训练基准测试v5.0结果显示，NVIDIA凭借Blackwell和Hopper架构继续占据主导地位，戴尔、超微等合作伙伴提交了相关系统；
➁ AMD展示了Instinct MI325X的显著性能提升，较前代MI300X提高30%，并在Llama2 70B训练中表现优于NVIDIA H200；
➂ 其他厂商参与度有限，谷歌Trillium首度亮相，反映出硬件性能提升与软件优化并行的行业趋势。
➀ MLPerf Training v5.0 results highlight NVIDIA's dominance with Blackwell and Hopper architectures, supported by submissions from Dell and Supermicro;
➁ AMD showcases improved performance of Instinct MI325X, claiming 30% generational gains over MI300X and competitive results against NVIDIA H200 in Llama2 70B training;
➂ Limited participation from other vendors noted, with Google's Trillium system appearing and a focus on accelerated hardware-software optimization trends.
---
本文由大语言模型（LLM）生成，旨在为读者提供半导体新闻内容的知识扩展（Beta）。
