Oracle will offer Google’s TPU AI chips via OCI in multi-gigawatt deployment
Will be the first hyperscaler to offer Google’s TPU outside of Google
Will be the first hyperscaler to offer Google’s TPU outside of Google
Project North Star will finance the purchase of Broadcom-developed Google TPUs
Company tells customers to pool mining capacity elsewhere
Agreement centers around Meta’s Muse AI agents
Facility will support electronic design automation (EDA), simulation, and offer sovereign AI cloud services
Could form first part of physical AI factory network
System will be powered by Nvidia GB200 Superchips and Vera CPUs
B300s and RTX gaming cards uncovered via exposed endpoints
The system will serve as the foundation of university’s quantum-RISE program
85 petaflops system features Nvidia H200s, housed in a repurposed building
Lab to host Nvidia and AMD hardware
Compute module to be placed on Astrolab rover
Has already been used to train Muse
The agreement will see the Department of Commerce take a minority stake in the company
The project expands the collaboration between Telefónica and IQM, which have already supplied two quantum computers to the Galicia Supercomputing Center (CESGA)
Rack scale solution can be deployed alone or alongside Nvidia Vera Rubin NVL72
Volatile market blamed for pause in public offering plans after months of investor qualms
Expanding support for agentic AI
All have been purpose-built for agentic AI, Dell said
Chris Malone announced new role two months after OpenAI departure was reported
Accessible via the Google Cloud A5X instances starting next month
Will start with Atlas systems and later adopt Titan
Says it will be the country's largest AMD Instinct deployment
‘Nvidia, clean up your dirty business,’ - protesters heckle Jensen Huang