Stony Brook University deploys Nvidia-based NVwulf cluster to support AI and ML workloads
System offers 80 petaflops of FP8 compute performance in its first phase
System offers 80 petaflops of FP8 compute performance in its first phase
Kvandal site will start at 230MW, could grow
Unclear if this builds on previous $4.3bn commitment
But lawmakers still unhappy with the decision to grant export licenses to China
Powered by 384 of its Ascend 910C chips, Huawei’s CloudMatrix servers offer more memory and bandwidth than Nvidia’s GB200 NVL72 at the cost of considerably more power
Cluster used by company’s SB Intuitions subsidiary now contains more than 10,000 GPUs
Comments made during Trump’s speech introducing US government’s new AI Action Plan
Comes as US lawmakers criticize move to grant export licenses
Built by HPE and powered by Nvidia GPUs, the system offers 23 exaflops of AI compute
Company’s Scope 1 and 3 emissions rise despite Scope 2 drops
Reversal follows meeting between President Trump and Nvidia CEO Jensen Huang
Instances now generally available
AI startup eyes investment from the Middle East
Chip firm continues to cash in on the AI revolution
Hosted by Hydra Host
Will support development of Emerald Conductor platform
Also this month, Chai AI announces $20m expansion of its in-house compute cluster
Nvidia GB200 NVL72 platform built by HPE coming to former Sharp LCD factory
Introduces new generation of Private Cloud AI, and announces collaboration with KDDI
The Karman platform was developed in collaboration with Nvidia
While Masayoshi Son hopes to build trillion-dollar robot and chip factory campus in Arizona
Another chip-related controversy
And other Scaleway updates
Expects to receive regulatory approval for the Natrium reactor in 2026