Nobel Laureate Michael Levitt Uses DeepSeek and Kimi Daily, Encourages AI Exploration

Chinese AI Startup DeepSeek Unveils Open-Source Optical Compression Model for LLM Training

This demonstrates that compact language models can effectively decode compressed visual text, potentially enabling larger models to adopt similar capabilities with fewer resources. The model is also highly scalable: a single A100-40G GPU can reportedly generate more than 200,000 pages of training data per day.

From Nvidia to DeepSeek: How AI Chips Are Shaping the U.S.-China Tech Rivalry

Amidst the ongoing global tech race and the shadow of geopolitical conflicts, if the United States continues to restrict the export of high-end AI chips, how can China challenge Nvidia's dominance in the AI chip sector?

DeepSeek V3.2 and GLM4.6 Large Models Set for Release

DeepSeek Updates Online Model to V3.1-Terminus With Improved Stability

China’s AI and Robotics Start-Ups Shine on Fortune’s Top 50 Tech List

DeepSeek Unveils V3.1 Model With Precision Upgrade for Next-Gen Domestic Chips

DeepSeek Posts Jobs on LinkedIn in Push to Attract Overseas AI Talent

Germany Asks Apple and Google to Remove China’s DeepSeek App Over Data Concerns

China's DeepSeek Upgrades Flagship AI Model R1

DeepSeek's announcement came just hours before Nvidia, the dominant U.S. supplier of AI chips, reported quarterly earnings.