hybrid inference model

DeepSeek’s New Breakthrough! FP8 Ignites the Revolution in Domestic Chips

by
DeepSeek's New Breakthrough! FP8 Ignites the Revolution in Domestic Chips

Overview On August 21, Chinese AI company DeepSeek quietly released its new large model DeepSeek-V3.1, which for the first time adopts a domestically developed precision standard called “UE8M0 FP8,” specifically tailored for the next generation of autonomous chips. This move not only marks the realization of “soft and hard synergy” in Chinese AI but also … Read more

Homo Intelligent NPU First to Achieve Edge Deployment of Alibaba Qwen3 Series Models

by
Homo Intelligent NPU First to Achieve Edge Deployment of Alibaba Qwen3 Series Models

On April 29, Alibaba Cloud launched the Qwen3 series of open-source hybrid inference models.In less than a day, Homo Intelligent’s self-developed NPU quickly achieved efficient deployment of the Qwen3 series models (Qwen3 0.6B-14B) at the edge.This achievement fully demonstrates the significant advantages of Homo Intelligent’s NPU in ecological adaptability and rapid response capability. Operation diagram … Read more