Nvidia happy to let Intel produce its future GPUs
WTF?! Could your future Nvidia GPU be produced by an Intel foundry? As unlikely as that sounds, team green boss Jensen Huang says it’s something his company is considering. Intel boss Pat Gelsinger responded to the news by saying he was “thrilled” for Nvidia’s interest, confirming that the two companies were engaged in ongoing discussions.
Huang answered questions at a press Q&A session yesterday (via Tom’s Hardware) and was asked about Intel’s Foundry Services (IFS) initiative. “We are very open-minded to considering Intel, and I’m delighted by the efforts that they’re making,” he said.
Intel and Nvidia compete in the CPU And GPU markets so the revelation may seem a little strange, but Haung explained that “Our strategy is to expand our supply base with diversity and redundancy at every single layer. At the chip layer, at the substrate layer, the system layer, at every single layer. We’ve diversified the number of nodes, we’ve diversified the number of foundries, and Intel is an excellent partner of ours […] They’re interested in us using their foundries, and we’re very interested in exploring it.”
Intel CEO Pat Gelsinger told Reuters that the company is “thrilled for their [Nvidia] interest in using our foundry capabilities.” He confirmed that discussions were ongoing, but there were no solid timelines—Huang added that foundry talks can take a long time.
“it’s not just about desire,” Huang said. “We have to align technology, the business models have to be aligned, the capacity has to be aligned, the operations process and the nature of the two companies have to be aligned. It takes a fair amount of time and a lot of deep, deep discussion – we’re not buying milk here. This is really about the integration of the supply chains. Our partnerships with TSMC and Samsung in the last several years are something that took years to cultivate. So we are very open-minded to considering Intel, and I’m delighted by the efforts that they’re making.”
Most companies would be hesitant to work so closely with competitors in case it exposes trade secrets, but Huang isn’t worried: he says both Intel and AMD have known Nvidia’s secrets for years, and there are plenty of legal ways to keep behavior on the level, such as NDAs. “We share roadmaps, of course, under confidentiality and a very selective channel of communications. The industry has just learned how to work in that way.”
Speaking about tech competitors collaborating in mutually beneficial ways, Huang cited some examples of how his company has worked with AMD and Intel, including the former’s CPUs appearing in DGX systems. “If not for Intel CPUs and all of the hyperscalers connected to our HGX, we wouldn’t be able to ship HGX,” Huang said, adding that Intel processors are used in Nvidia’s Omniverse platform.
Nvidia is currently focused on producing this year’s RTX 4000 series, aka Lovelace, for which it has reportedly paid TSMC $10 billion to secure a significant chunk of the Taiwanese firm’s 5nm manufacturing capacity. The company currently uses Samsung’s 8nm process node to make the consumer Ampere GPU line.