Advancements in semiconductor technology, GPU development, and data center infrastructure that power AI computing capabilities.
AstraZeneca and Oracle will combine clinical expertise, health data, cloud infrastructure, and AI to support trial recruitment, diagnosis, safety monitoring, and care workflows.
CBRE agrees to acquire data center infrastructure provider Onnec for an initial £460 million in cash, plus an earnout of up to £300 million.
A confidential client is exploring buying the closed San Francisco Centre mall to turn its upper floors into a data center and AI lab, as the city pauses data center permits.
Elice shares fell as much as 33% on their first day of trading after the AI infrastructure company raised 201 billion won in its IPO.
OpenAI has stopped pressing lawmakers for new chip export limits on China, aides and lobbyists told Politico, a shift that helps Nvidia.
G42, AMI Labs and Analog signed nonbinding agreements covering world model research, AI infrastructure, cloud services and possible industrial applications.
Google will cover new power resources and grid upgrades for its expanded West Memphis data center campus. Entergy Arkansas projects at least $500 million in additional customer savings over 20 years.
DeepSeek plans to invest in Huawei's AI chip design and help improve Ascend processors for training large language models.
NVIDIA has released the full Run:ai platform as open source software, including four modular tools for scheduling workloads and managing GPU resources on Kubernetes.
Amazon Bedrock now supports structured outputs for all available Anthropic Claude models, providing JSON responses and tool parameters that conform to developer defined schemas.
Greece has inaugurated Daedalus, a €58.9 million supercomputer built to support AI applications and scientific research. The system delivers measured performance of 85.69 petaflops and ranks 31st on the TOP500 list.
The US imposed sanctions on Shanghai Zhisuan Technology, MiraclePlus and three related entities following the conviction of former Google engineer Ding Linwei for stealing AI trade secrets.
Anthropic and Nscale are reportedly nearing another $45 billion compute agreement after signing a deal of the same size in August. Discussions remain underway, and the terms could change.
S&P Global says major technology companies may have more liabilities tied to AI infrastructure than public disclosures show, but it does not currently plan to change their credit ratings.
Nova Scotia opposition parties want an emergency legislative sitting to examine consultation, regulation and power costs for EverWind Fuels' planned Anthropic data centre.
Deepgram ported its Nova-3 speech recognition model to AWS Trainium2 for offline batch workloads. The resulting setup processed 11% to 70% more audio per dollar than its previous compute options.
A Vanderbilt Policy Accelerator report proposes using the Defense Production Act to reserve scarce memory chips for industries outside AI.
Modal plans to deploy Nvidia Vera CPUs for AI agents after tests showed they completed twice as much work as its current processors.
Wisconsin regulators will conduct a statewide environmental review of AI data center growth under a settlement over permits for the $15 billion Lighthouse campus.
TorchNeuron lets developers run existing PyTorch code on AWS Trainium chips with a configuration change instead of a rewrite.
Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.