最新文章

共 31 篇

DeepSeek-V3.1混合推理架构解析：开启智能体时代新篇章

DeepSeek-V3.1 introduces hybrid inference with Think/Non-Think modes, enhancing reasoning efficiency and agent capabilities while supporting 128K context and updated APIs. (DeepSeek-V3.1引入思考/非思考混合推理模式，提升推理效率和智能体能力，支持128K上下文和更新的API。)

DeepSeek2026/1/22

阅读全文 →

DeepSeek-R1推理模型发布：性能媲美OpenAI-o1，开源助力AI研究

暂无摘要...

DeepSeek2026/1/22

阅读全文 →

AAEON发布三款NVIDIA驱动AI边缘计算设备：紧凑设计赋能高效部署

AAEON has launched three NVIDIA-powered edge AI devices—BOXER-8251AI, BOXER-8240AI, and BOXER-8222AI—featuring compact designs, ARM processors, and CUDA cores for efficient edge computing. (AAEON推出了三款NVIDIA驱动的边缘AI设备——BOXER-8251AI、BOXER-8240AI和BOXER-8222AI，采用紧凑设计、ARM处理器和CUDA核心，实现高效的边缘计算。)

AI大模型2026/1/21

阅读全文 →

内存计算(CIM)架构突破：2024年解决AI能效危机指南

Memory computing (CIM) addresses AI's energy crisis by processing data where it's stored, reducing data movement energy by up to 80% and enabling sustainable zettascale computing. (内存计算(CIM)通过在数据存储位置直接处理数据解决AI能源危机，减少高达80%的数据移动能耗，实现可持续的泽塔级计算。)

AI大模型2026/1/21

阅读全文 →

AI推理框架2024指南：从理论模型到生产应用关键技术解析

AI inference frameworks are specialized software systems that execute trained machine learning models in production environments, optimizing for performance, efficiency, and scalability through techniques like quantization, pruning, and hardware acceleration. (AI推理框架是专门在生产环境中执行训练好的机器学习模型的软件系统，通过量化、剪枝和硬件加速等技术优化性能、效率和可扩展性。)

AI大模型2026/1/21

阅读全文 →

AI推理框架工业应用指南：高效自动化流水线2024

AI inference frameworks are high-efficiency information automation pipelines based on statistical patterns and interpolation, excelling in industrial applications but limited in creativity. (AI推理框架是基于统计规律和插值计算的高效信息自动化流水线，在工业应用中表现出色，但在创造性方面存在局限。)

AI大模型2026/1/21

阅读全文 →