Leading developers of AI models from China want Nvidia's Rubin and explore ways to rent the upcoming GPUs in the cloud.

A new report says Nvidia's presence is about to shrink drastically in China.

U.S. introduces new export rules for AI and HPC accelerators developed in America that allow shipments of AMD's MI325X and Nvidia's H200 GPUs to Chinese entities, but only if their developers meet demand of American customers and meet other requirements of the Department of Commerce.

GlobalFoundries acquires more CPU IP as it is reshaping its role in the semiconductor supply chain and becomes an IP and manufacturing partner rather than just a chipmaker.

TSMC executive team expects continuous revenue growth in 2026, and onwards as rich AI companies will keep buying AI accelerators for years to come.

SK hynix is investing $12.9 billion to build a campus-scale, HBM-only advanced packaging and test facility in Cheongju, South Korea, designed for the next generation of HBM memory and intended to ensure SK hynix's leadership in the booming market.

The U.S. may allow shipments of rather powerful AI processors to China on a case-by-case basis, but with the U.S. supply priority, do not expect AMD or Nvidia ship a ton of AI GPUs to the People's Republic.

Phison's aiDAPTIV+ stack enables large Mixture of Experts AI models and agentic AI workloads to run on client systems with limited memory capacity.

SK Hynix demonstrates 48 GB HBM4 memory with a 2,048-bit interface over at up to10 GT/s

Meta establishes 'Meta Compute' organization to manage buildout of next-generation data centers and development of custom computing architectures that will power gigawatt-class data centers.

As the AI hardware market is about to change, Nvidia is hiring its first chief marketing officer.

Intel's CEO says he expects 'great momentum in terms of yields and IP' on the company's 14A process technology.

Chinese researchers have demonstrated a 4F² dual-gate 2T0C capacitor-less, DRAM-like memory cell with multi-bit storage, fast writes, and long retention. In theory, the technology could be used as embedded or stacked 3D memory, but questions remain about its manufacturability and commercial viability.

Sandisk and other key suppliers of 3D NAND are projected to significantly increase prices of enterprise-grade 3D NAND memory in the coming months due to overwhelming demand from the AI sector.

Nvidia has reportedly imposed non-refundable, full upfront payment terms on H200 GPU sales to China-based clients to hedge regulatory risks as Chinese customers had ordered more than 2 million H200 units that the company yet has to make.

After numerous delays, Micron is about to start building its $100 billion site in New York to produce 40% of its DRAM output in the U.S. in the 2040s.

Elon Musk says that modern cleanrooms are built wrong, and if Tesla builds its own fab, he will be able to eat and smoke in that facility.

AMD announces CDNA 6-based Instinct MI500-series GPUs with HBM4E memory set to arrive in 2027 and offer 1000X higher performance compared to Instinct MI300X.

AMD reveals additional details about its strategy for AI and HPC deployments this year with EPYC 'Venice' and 'Venice-X' CPUs as well as Instinct MI430X, MI440X, and MI455X accelerators.

Cables and pipelines in the Gulf of Finland have been damaged four times in less than 1.5 years.