The rise of agentic AI is fueling fresh debate over the future GPU-to-CPU balance in AI systems, with Arm CEO Rene Haas now weighing in on the discussion. According to a transcript published by Investing.com, Haas said that while CPUs may not outnumber GPUs on a chip basis, they could from a core-co...

As Beijing continues pushing for greater semiconductor self-sufficiency, Chinese chip companies are allocating a larger share of revenue to R&D. According to South China Morning Post, first-quarter earnings show that China’s leading chipmakers devoted a higher proportion of revenue to research...

AI-driven demand is surging across the supply chain, with copper clad laminates (CCL) also coming under significant strain. According to Hankyung, citing sources, a Seoul-based PCB maker placed advance orders worth KRW 10 billion with Taiwan’s EMC and TUC—more than five times its typical monthly...

While Google has signaled a strategic shift to offer its Tensor Processing Units (TPUs) to select external customers—moving beyond its legacy internal-only model—market reception among leading AI infrastructure providers remains cautious. Citing The Information, the Chosun Daily reports that spe...

Intel is collaborating with SoftBank’s subsidiary SAIMEMORY to advance a potential alternative to conventional high-bandwidth memory (HBM). According to TechPowerUp, SAIMEMORY plans to present a paper at VLSI Symposium 2026 in June, detailing its newly developed HB3DM memory based on Z-Angle Memor...