China's LineShine supercomputer, built by the Shenzhen Cloud Computing Center, has become the world's fastest, overtaking the US's El Capitan. This marks the first time since 2017 that China holds the top spot, with LineShine performing 20% better on the High Performance Linpack benchmark. Unlike most top supercomputers, LineShine runs on over 13 million CPUs rather than GPUs. Despite this, the US still dominates the overall rankings, holding the next three fastest spots. This development comes as the US ramps up efforts in quantum computing, highlighted by a recent executive order from President Donald Trump.
TSMC and ASML have both increased their guidance, indicating ongoing supply constraints in AI infrastructure. Nvidia's Rubin N3 node is fully booked, and CoWoS capacity has expanded by nearly 50%, supporting strong demand. Concerns about Nvidia's Kyb...
Nvidia introduced Cosmos 3 Edge, a new AI model designed for robots and vision AI agents to navigate physical environments in real time. The company is expanding its AI ecosystem in Japan by partnering with major local firms like Fujitsu, Hitachi, an...
Netlist has prompted the US International Trade Commission (ITC) to investigate Samsung and its customers Google, Supermicro, Nvidia, and Broadcom for allegedly infringing two Netlist patents related to Samsung's high-bandwidth memory and DDR5 memory...
NVIDIA and Noetra Corp. are partnering to build Japan's first national AI infrastructure, the NVIDIA Vera Rubin AI factory, featuring 13,750 Vera CPUs and 27,500 Rubin GPUs. Supported by Japan's Ministry of Economy, Trade and Industry, this 140-megaw...
NVIDIA announced that leading Japanese companies in robotics and manufacturing are adopting its Cosmos, Isaac, Metropolis, and Jetson platforms to accelerate AI-powered intelligent machines across industries. The new Cosmos 3 Edge model enables real-...
