Dell CEO Michael Dell announced the company is building a massive AI factory supercomputer leveraging Nvidia’s latest and greatest GPUs for the express purpose of powering Elon xAI outfit and its Grok language model.
In a post on X, Dell revealed the Dell AI Factory collaboration with Nvidia to provide the heavy-duty compute muscle for training future iterations of xAI’s flagship Grok model.
Elon has made it crystal clear that xAI is going all-in on an AI training arms race, claiming the 100k Nvidia H100 tensor core GPUs to reach its full potential, (Grok-1.5V Brings Vivid Visuals to the Multimodal AI Revolution). For context, that’s at least four times larger than the biggest existing GPU clusters on the planet.
Elon has already been snatching up these $40,000 accelerators by the tens of thousands over the past year for prior Grok model training. New Dell/Nvidia partnership looks primed to take xAI’s hardware hunger to new heights for that all-important OpenAI competitor.
Of course, details are still scarce on the scale and specs of AI Factory monster rig Dell is cooking up. Based on Elon’s voracious GPU appetite and the sheer computational requirements needed to take on AI juggernauts like OpenAI, we’re likely talking about one of the most powerful supercomputers ever conceived.
The big question looming? Whether Elon’s money pit of an AI moonshot can possibly match the technical prowess and cost-effectiveness of purpose-built cloud training solutions offered by Amazon, Microsoft, and Google. Dell/Nvidia power play is undoubtedly a major statement of intent, though.