Elon Musk's SpaceX set to produce 'own GPUs' at its own multi-billion fab as the company warns that it may be unable to purchase all the silicon it needs to meet its goals.
TSMC strengthens its bifurcated process technology development approach with A14, A13, and N2U aimed at client applications and A16, A12, and N2X for high-performance data center designs.
Cerebras' sales hit half a billion dollars in 2025, but 86% of its revenue comes from Abu Dhabi-based G42 and Mohamed bin Zayed University of Artificial Intelligence, which represents major risks.
By now, TSMC has become a much bigger chipmaker than Intel has ever been, but the world's top foundry still calls its American peer a 'formidable' competitor.
The U.S. Court of Appeals determines $3 billion VLSI lawsuit was not suitable for summary judgment and must instead be evaluated by a jury, so it must return to a district court for proceedings.
Ex-Apple, ex-Nuvia CPU designers Gerard Williams, John Bruno, and Ram Srinivasan establish Nuvacore startup to develop clean-sheet CPUs for always-on compute-intensive AI workloads.
Nvidia widely uses AI for different stages of the chip design process, though it admits that AI is a long way from designing chips completely autonomously.
Although Google now has its own Arm-based Axion CPUs, Intel's Xeon processors with custom IPUs will continue to be used for AI and other demanding workloads in Google's data centers for years.
Hacker or hacker group steals secret data concerning aerospace engineering, bioinformatics, fusion modeling from China's National Supercomputing Center.
We take a deep-dive into the past, present, and future of the ubiquitous PCIe standard, and look ahead at the challenges that await manufacturers when integrating PCIe 6.0 and beyond into real-world hardware.
Anmelden
