Search AI Products and News

DeepSeek launched an official account on Zhihu and published a technical article, "DeepSeek-V3/R1 Inference System Overview," detailing its model inference system optimization and cost-profit margin information for the first time, marking the end of the highly anticipated DeepSeek Open Source Week. The article highlights two key optimization goals for the DeepSeek-V3/R1 inference system: higher throughput and lower latency. To achieve these goals, DeepSeek employs large-scale cross-node expert parallelism (EP) technology, although...

DeepSeek announced its second-day product of the open-source week: the first open-source EP communication library for MoE models, enabling full-stack optimization for mixed-expert model training and inference. DeepEP is a high-efficiency communication library specifically designed for Mixture-of-Experts (MoE) and Expert Parallelism (EP). It aims to provide high-throughput and low-latency many-to-many GPU kernel communication, commonly known as MoE routing and aggregation. DeepEP not only supports low-precision operations such as FP8 but also integrates with the DeepSeek-V3 paper.