Nvidia, Oracle, and the U.S. Department of Energy will build seven ExaFLOPS-class AI supercomputers for Argonne National Laboratory — including the Oracle-built Equinox and Solstice systems with over 100,000 Blackwell GPUs delivering up to 2,200 FP4 ExaFLOPS — to power next-generation AI and scientific research.

Nvidia introduces Omniverse DSX Blueprint, a digital-twin-based reference design for gigawatt-scale AI data centers that standardizes how partners can build and optimize 'AI factories.'

Elon Musk's statement that Samsung's Taylor, Texas fab is more advanced than TSMC's Fab 21 in Arizona reflects the newer 3nm-era tools being installed there. However, this advantage has little relevance for Tesla's AI5 processor, which likely relies on SF4A FinFET technology, which gains minimal benefit from those capabilities.

Nvidia and HPE will build the Mission and Vision supercomputers for Los Alamos based on the Vera Rubin platform to advance national-security and open-science research using AI simulation and scientific computing.

SK hynix introduced its AI NAND lineup — AIN D, AIN P, and AIN B — at the 2025 Global Summit, outlining a new strategy to deliver high-density, high-performance, and high-bandwidth storage tailored for AI servers and workloads.

Qualcomm has unveiled its AI200 and AI250 rack-scale AI inference solutions relying on data center-grade Hexagon NPUs with near-memory computing, micro-tile inferencing, and confidential computing support.

Intel cut 20,500 jobs under new CEO Lip-Bu Tan — totaling 35,500 in two years — and reduced R&D spending by over $800 million per quarter as it streamlines operations, cancels lower-priority projects, and concentrates investment on strategic projects.

Intel's third-quarter results were constrained by production and substrate shortages, forcing the company to prioritize data center CPUs over client chips, which will likely make Raptor Lake CPUs more expensive going forward.

AMD is investing over $280 million to establish two new R&D centers in Taiwan dedicated to silicon photonics and heterogeneous integration to further boost its data center roadmap.

Intel said that its 18A process is advancing on schedule with usable yields to begin Panther Lake volume production, but low marings will slow the ramp and capacity expansion until yields reach 'industry standard' levels in 2026 – 2027.

Intel returned to profitability in Q3 2025 with $13.7 billion in revenue and $4.1 billion profit, driven mainly by one-time gains, as client and data-center sales improved despite ongoing capacity shortages.

A Chinese enthusiast’s benchmark suggests Apple’s new M5 chip delivers record-breaking single-thread performance in CPU-Z, outperforming top AMD and Intel CPUs even while running Windows 11 in virtualization mode. But its multi-thread performance should be better.

Japan's Automobile Manufacturers Association warned that Nexperia's disrupted chip supply could halt vehicle production across Japan, as Volkswagen is reportedly preparing to suspend assembly of Golf and Tiguan models as early as next week.

The U.S. government is in talks to take equity stakes in major quantum computing startups in exchange for at least $10 million each in funding from the Commerce Department.