今日DeepSeek官宣,今日开源DeepEP,第一个用于 MoE 模型训练和推理的开源 EP 通信库。特点有:高效、优化的全员沟通;节点内和节点间均支持 NVLink 和 RDMA;用于训练和推理预填充的高吞吐量内核;用于推理解码的低延迟内核;原生 FP8 调度支持;灵活的 GPU 资源控制,实现计算-通信重叠。风险提示及免责条款 市场有风险,投资需谨慎。本文不构成个人...
Source LinkDisclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.