Intel wants to make chips for NVIDIA, AMD, Qualcomm, Google, and Microsoft. America’s largest semiconductor foundry wants clients from every niche and segment of the industry, from all over the world. Company CEO, Pat Gelsinger stated as much during Intel’s foundry event the other day. After years of lagging behind its Asian rivals, Intel is separating its design and foundry divisions in all but name (maybe that too).
Gelsinger claims that Intel will soon post separate financials for the design and foundry divisions. There will be a separate legal entity for Intel Foundry, and its sole objective will be to fill its manufacturing capability to the brim. There’s no discrimination here either. Intel’s fabs are welcoming its archrivals AMD, Arm, NVIDIA, Qualcomm and more. The Arm CEO being present at the event says a lot about how serious Gelsinger is.
The foundry team’s objective is simple: Fill. The. Fabs. Deliver to the broadest set of customers on the planet. We hope that that includes Jensen (Nvidia), Christiano (Qualcomm), and Sundar (Google), and you heard today it includes Satya (Microsoft), and I even hope that includes Lisa (AMD) going forward.
Pat Gelsinger
Arm-based Neoverse processors are already been designed at Intel’s cutting-edge 20A/18A process nodes, and it’s only a matter of time before they’re taped out. The Intel CEO went as far as to name the CEOs he’d like to work with, and it’s a long list. Notable names include Jensen Huang (NVIDIA), Christiano Amon (Qualcomm), and of course Dr. Lisa Su (AMD).
Intel isn’t shying away from custom designs either. The advanced packaging technologies used in Clearwater Forest, including Foveros Direct, EMIB, and copper-copper hybrid bonding will also be made available to third-party clients. This will go a long way in solving the packaging capacity shortage faced by NVIDIA and AMD at TSMC’s CoWoS foundries.
How to do hybrid bonding, Intel 3 base die, Intel 18A top die, being able to solve a lot of the CoWoS/Foveros problems using EMIB and hybrid bonding. That will become a set of collaterals that will benefit the foundry team.
Pat Gelsinger
If the Intel Foundry is a separate legal entity with its own budget, expenditure, and financials, just how far is it from completely spinning off as a separate company like Global Foundries back in the day? I believe it’s only a matter of time before that happens. If Intel wants its archrivals (AMD, Arm, and NVIDIA) to fully commit to its American fabs, it’ll take more than a few keynotes and slideshows to cement that trust.
There are Intel products and Intel foundry, There’s a clean line between those, and as I said on the last earnings call, we’ll have a setup separate legal entity for Intel foundry this year. We’ll start posting separate financials associated with that going forward.
Pat Gelsinger
There’s a reason why TSMC is the leading fab, and not Samsung or Intel. It’s completely neutral (even though it has its preferred clients) with a history of delivering on its roadmap. Intel has neither. Even if we ignore traditional rivalries, how will AMD’s shareholders react to an Intel Foundry partnership? What if Intel fails to deliver on the promised capacity or delays its orders?
For these reasons, it’s unlikely that AMD or NVIDIA will commit a large chunk of their orders to Intel Foundry, at least not anytime soon. If initial orders are delivered timely, without unexpected defects or delays, then this strange setup may go places. However, the uncertainties surrounding it are just too many at the moment.
If Pat’s comments are to be taken at face value, then I believe Intel has already started the division of its design and foundry businesses. First, the financials and the legal, then the branding, the leadership, and soon the rest.
Well, if you go back to the picture I showed today, Paul, there are Intel products and Intel foundry, There’s a clean line between those, and as I said on the last earnings call, we’ll have a setup separate legal entity for Intel foundry this year. We’ll start posting separate financials associated with that going forward. And the foundry team’s objective is simple: Fill. The. Fabs. Deliver to the broadest set of customers on the planet.
We hope that that includes Jensen (Nvidia), Christiano (Qualcomm), and Sundar (Google), and you heard today it includes Satya (Microsoft), and I even hope that includes Lisa (AMD) going forward. I mean, we want to be the foundry for the world, and if we’re going to be the Western foundry at scale, we can’t be discriminating about who’s participating in that. So, unequivocally, it is to be the foundry for the world. Commit supply chains, your leadership technology – the doors to the ala carte menu are wide open for the industry.
And some of those [garbled] that we build, like Clearwater Forest that I showed today, hey, that’s a construct that was innovated by my Xeon team: How to do hybrid bonding, Intel 3 base die, Intel 18A top die, being able to solve a lot of the CoWoS/Foveros problems using EMIB and hybrid bonding. That will become a set of collaterals that will benefit the foundry team. They’re going to go sell that constructional opportunity as a better way to build AI chips. So clearly, I’m taking product group intellectual property and leveraging it on the foundry side of the business.
So I want my foundry to be used by everybody. Period. We want to help build Nvidia chips, and AMD chips, and TPU chips for Google, and inference chips for Amazon. Period. We want to help them and give them the most powerful, performant, and efficient technologies for them to build their systems. Period. Full stop.
Intel CEO to Tom’s Hardware
