Nvidia's Delayed Blackwell AI Chips Overheating in Servers
The server rack Nvidia designed for Blackwell—which can connect up to 72 GPUs at a time—is reportedly causing the overheating issue. Nvidia has repeatedly redesigned the racks, which could result in GPU server shipments being delayed and new Google ...
Nvidia redesigns 72-GPU AI server racks after Blackwell GPUs overheat - report
Speaking to DCD earlier this year, Ian Buck, Nvidia’s accelerated computing VP, said: "The fabric of NVLink, the spine, is connecting all those 72 GPUs to deliver an overall performance of 720 petaflops of training, 1.4 exaflops of inference."
When fully loaded, the 72-GPU GB200-based NVL72 rack weighs ... and a new 4-Way GB200 (Grace-Blackwell) boards that still ...
Blackwell is Nvidia's new flagship graphics processing unit architecture ... The engineering iterations are normal and ...
Production of the popular graphics card is reportedly being wound up to make way for the new RTX 5000 series, including the ...
Traditionally, Nvidia's stock surges after its earnings report and is already up 8% this month. Analysts predict over a ...
Results that may be inaccessible to you are currently showing.
Hide inaccessible results
Feedback