In a recent episode of the Bg2 Pod, Nvidia Corporation’s CEO Jensen Huang shared his insights on various topics, including Elon Musk’s xAI, the artificial intelligence company founded by the Tesla and SpaceX CEO. During a conversation with Altimeter Capital’s CEO Brad Gerstner and partner Clark Tang, Huang was asked about xAI’s remarkable accomplishment of constructing a large, coherent supercluster in Memphis in a mere matter of months.
Huang expressed his deep respect for Musk’s capabilities, stating, “Elon is singular in this understanding of engineering and construction and large systems and marshaling resources.” He went on to commend the engineering, networking, and infrastructure teams at xAI, praising the “incredible” integration of technology and software.
Huang provided a compelling perspective on the scale of this achievement: “Just to put in perspective, 100,000 GPUs that’s you know easily the fastest supercomputer on the planet as one cluster. A supercomputer that you would build would take normally three years to plan. And then they deliver the equipment and it takes one year to get it all working.” He then added, “We’re talking about 19 days.”
This astonishing speed underscores the dedication and expertise of the xAI team. Earlier this year in July, xAI initiated the training of the Memphis Supercluster, equipped with 100,000 Nvidia H100 GPUs, solidifying its position as the world’s most powerful AI training cluster. Reports had surfaced earlier of Musk and Oracle’s Larry Ellison seeking additional GPUs from Huang during a dinner meeting. This discussion on the Bg2 Pod further highlights the strong relationship between Musk and Huang, a relationship evident in Musk’s earlier praise for Huang’s work ethic.
Huang has also previously expressed his appreciation for Musk’s endeavors, particularly in the domain of self-driving vehicles. This mutual admiration between two prominent figures in the technology world speaks volumes about their commitment to advancing innovation and pushing the boundaries of what’s possible.
