Search results for GB200
Industry
AI factory build in Korea leaves server, cooling supply chain tied to Taiwan
Construction of gigawatt-scale AI factories in South Korea is getting under way, but server assembly and liquid-cooling component supply chains are controlled by Taiwanese companies, raising concerns that Korean materials and parts suppliers are being sidelined. Nvidia plans to ship fully assembled compute trays from its next-generation Vera Rubin platform, narrowing assembly partners to Foxconn, Quanta and Wistron. Cooling component suppliers Auras and AVC have also taken an early lead.
AI & Enterprise
DeepSeek founder says China can end Nvidia dependence within a year
DeepSeek founder Liang Wenfeng said China’s AI industry is reaching a turning point in its dependence on Nvidia hardware and that chip supply will no longer be a key obstacle within a year. He cited a resource gap with the United States driven by computing power constraints. He said DeepSeek lags U.S. leaders by 12 to 18 months but achieves similar performance with far less computing. DeepSeek is working more closely with Huawei and pursuing its own chips and software tools.
Games & Commerce
Naver pivots from search to AI infrastructure with \'AI factory\' push
Naver is positioning AI infrastructure as a new growth pillar, moving beyond domestic search, shopping and advertising into B2B infrastructure aimed at global AI computing demand. Its core project is an “AI factory” 추진 with Nvidia. Plans call for ramping capacity from 55MW in the first half of next year to 200MW by end-2028, with a long-term goal of 1GW. Key hurdles include long-term customer contracts, funding, power and sites, and technology-change risks.
-
AI & Enterprise
Tesla moves into modular AI data centre hardware with Megapod trademark filing
-
AI & Enterprise
Nvidia H200 GPU rental rates fall 40 percent in three weeks, raising questions over AI chip scarcity
-
AI & Enterprise
Nvidia supplies GPT-5.5 Codex to 10,000 employees, cutting cost to one-35th
-
AI & Enterprise
Vesl AI secures GB200 and B300, says only domestic GPU cloud provider with both
-
AI & Enterprise
Google in talks to invest $100 million in Fluidstack, boosting TPU ecosystem