MiniMax (official)· @MiniMax_AI · X·· 2026-07-23AI 评分51
AI 导读
MiniMax 宣布 AMD MI355X 在其 M3 模型推理上已接近 Nvidia B200 水平。双方联合发布技术博客，详述针对 428B MSA 稀疏 MoE 多模态基座模型的全栈 ATOM + ATOMesh 协同优化，核心升级包括 EAGLE3 投机解码、支持 1M 超长上下文的 MSA 对齐 Page-16 KV 缓存、逐层 FP8 量化及分布式 P/D 分离架构。
正文 · 原文
We welcome more hardware manufacturers to join us in jointly optimizing inference performance, OSS forever!
AMD MI355X now is close to Nvidia B200 on MiniMax-M3 serving. Thanks to @AMD team. released dual technical blogs detailing full-stack ATOM + ATOMesh co-optimization for our 428B MSA sparse MoE multimodal base model. • EAGLE3 speculative decoding for lower generation latencyCore upgrades: • MSA-aligned Page-16 KV cache for 1M ultra-long context • Layer-wise FP8 quantization avoiding vision/MoE precision drop • Distributed P/D separation architecture for large-scale clusters 🔗 Blog 1: http://rocm.blogs.amd.com/artificial-intelligence/minmax-m3-atomesh/README.html 🔗 Blog 2: http://rocm.blogs.amd.com/software-tools-optimization/minimax-m3-mi355/README.html在 X 查看被引用的帖子
来源：MiniMax (official) · x.com