Realtime AI News
Anthropic in Talks With Samsung to Develop Custom AI Chip
Anthropic is negotiating with Samsung Electronics to co-develop a custom AI chip, aiming to reduce reliance on Nvidia GPUs. The partnership could reshape the AI hardware supply chain. Sources say discussions are at an advanced stage but no final agreement has been signed.
According to sources familiar with the matter, Anthropic and Samsung are in advanced negotiations to jointly design a custom AI chip optimized for large model training and inference. The chip would leverage Samsung's advanced fabrication processes and incorporate Anthropic's architectural expertise.
If the deal goes through, Anthropic would gain more tailored computing power while reducing its dependence on Nvidia GPUs. Samsung, in turn, would secure a marquee customer in the AI chip space, bolstering its position in the custom chip market.
The talks are ongoing and no final agreement has been reached. However, this move signals that top AI companies are actively seeking to diversify their hardware supply chains to mitigate single-vendor risk.
Notably, Anthropic already has a deep partnership with Google Cloud, while Samsung has strengths in both memory and foundry services. A successful collaboration could further shift the AI chip industry landscape.
Key follow-ups include the chip's specific architecture, production timeline, and its potential impact on the existing GPU market.
Why it matters
This collaboration, if realized, would accelerate the trend toward custom AI chips and potentially alter the competitive dynamics of AI hardware.
Nearby Updates
All07/03, 22:28
GPT-5.6 vs Claude Fable 5 Coding Benchmarks Reveal Diverging Performance
Latest coding benchmarks show GPT-5.6 leads in code generation tasks while Claude Fable 5 excels in code reasoning and debugging. The two models display distinct strengths, with GPT-5.6 outperforming by 12% on generation but trailing by 8% on reasoning tasks.
07/03, 15:35
Alibaba DAMO Academy Unveils AI Agent ElementsClaw, Discovers 4 New Superconductors in 28 GPU Hours
Alibaba DAMO Academy, together with Renmin University and the University of Chinese Academy of Sciences, released ElementsClaw, the first AI agent dedicated to superconductor discovery. The system screened 2.4 million stable crystal structures in just 28 GPU hours and discovered 4 previously unknown superconductors through experimental validation.
07/03, 13:24
China's BGI Subsidiary and Shanghai AI Lab Release ProtoPilot, First AI System to Complete Wet-Lab Experiments End-to-End, Outperforming GPT-5.6 Sol
Yongsheng Intelligence, a subsidiary of BGI, together with the Shanghai Artificial Intelligence Laboratory, released ProtoPilot and BioLab Bench, achieving the first full loop from natural language experimental intent to physical wet-lab execution. Third-party evaluations show it surpasses OpenAI's flagship GPT-5.6 Sol in end-to-end life science agent capabilities.
07/03, 11:36
WAIC 2026 to Spotlight Computing Power Breakthroughs: Super Nodes and Optical Interconnects
The World AI Conference 2026 (WAIC), running July 17-20, is putting computing infrastructure at center stage. The event will explore whether super-node architectures and optical interconnect technologies can bypass the physical limits of single-chip performance for AI workloads.