Sources: Nvidia is considering lower-memory versions of its Rubin Ultra GPU due to potential issues securing enough HBM, and has tested at least three versions (The Information)
Nvidia is weighing a radical step to deal with a shortage of advanced high-bandwidth memory chips: using less of it than planned …
Nvidia is considering lower-memory versions of its Rubin Ultra GPU due to potential issues securing enough high-bandwidth memory chips. According to The Information, the company has tested at least three versions of the GPU with reduced memory.
In other news, Elon Musk has stated that SpaceX and xAI will exclusively use Nvidia GPUs, citing their superiority. During SpaceX's earnings call, Musk praised Nvidia's Vera Rubin architecture, saying it is the best AI computer and that the design of the NVL72 rack-scale system is superior to others.
Musk also mentioned that xAI has been using Nvidia's accelerators, including Hopper and Blackwell hardware, to train multiple generations of Grok. There is no evidence to suggest that xAI has evaluated or used non-Nvidia accelerators in production.
Brief written by urgent.news from Techmeme, Tom's Hardware — 2 reports on this story. Machine-written — read the original for the full account.
We haven't written up this one. Techmeme has the full story — the link below goes straight to it.
This story
This is one outlet's version. Read the fullest account.
