Intel Foundry might be heading into a big deal with NVIDIA, as the company is expected to offer packaging and wafer support for next-gen Feynman GPUs.
NVIDIA Feynman GPUs Could Be Intel Foundry's Biggest Win To Date With Chipzilla Providing Both Packaging & Wafer Supply
A deal between Intel and NVIDIA won't be a major surprise for those following our previous reports, but it is definitely a big win for Intel's Foundry business. Earlier this year, we reported that NVIDIA might be looking into Intel's upcoming process & advanced packaging technologies for its next-generation GPUs, codenamed Feynman.
Well, according to recent chatter, it looks like this deal is moving forward, and Intel Foundry is indeed going to snag a packaging and wafer deal with NVIDIA for its future AI chip.
The latest indicator comes from @bubbleboi on X, who can be seen quoting Patrick Moorhead's (Founder/CEO of MoorInsStrat) post on Intel's recent earnings, and how they are increasing CapEx towards advanced packaging & fabs.
During the earnings call, Intel's CEO, Lip-Bu Tan, said the increased CapEx is a "signal of our confidence in customers across all of our business units". Under Lip-Bu's leadership, Intel has followed a disciplined approach to capital spending, as he has said that he won't put a dime towards things that aren't productive for the company. But now, Lip-Bu feels "very confident," and has teased "long-term agreements," and is now going all in to bring forth the capacity to meet the growing customer interest.
It's going to include advanced packaging. It's what we talked about. We're pretty excited about our prospects on eMIBT, and so we will be investing in that.
That said, the cost of a fab for the front end is much more expensive than a packaging facility, so it will be skewed towards the front end. But nevertheless, both of them will be important to us. As it relates to customers, I would just say that these increased investments is a signal of our confidence in customers across all of our business units.
We feel very confident, particularly in places where we've gotten long-term agreements, that we now have the signal to be able to kind of forecast out what the outlook looks like for the next few years in terms of demand, and we're putting forth the capacity in anticipation of that across all of our businesses.
Lip-Bu Tan - Intel CEO (Q2 2026 Earnings Call)
Lip-Bu has also said that he will honor the customer and won't name anyone. But it looks like one big customer has been hidden in plain sight, and that might indeed be NVIDIA.
NVIDIA + Intel - A Growing Bond
NVIDIA & Intel have had a much closer relationship under Lip-Bu. Both companies announced a deal where Intel will build custom Xeon chips with NVIDIA's NVLink interconnect solution and also feature NVIDIA's RTX GPUs on its client SoCs (these are called Serpent Lake internally). NVIDIA released an entry-tier Rubin NVL8 solution which houses Intel Xeon 6 as the host CPU, and so far, Intel and NVIDIA have seen a fruitful relationship together.
But Lip-Bu Tan is a crowd pleaser, and his decades-long experience and relations with the tech industry have already moved several of his friends to come and join him as new Intel leaders. His team continues to expand, and Lip-Bu is now prioritizing his Foundry plans for one of his biggest wins at Intel to date.
Intel Has The Right Tools at The Right Time For Feynman
As mentioned, Intel won't spend big if they don't have a big customer, and NVIDIA's Feynman is going to be massive. Feynman GPUs are interesting; they come after the Rubin generation around 2028, and will feature some big enhancements such as 3D Die-Stacking, a brand new HBM standard, and a new host CPU, called Rosa.
NVIDIA Data Center / AI GPU Roadmap
| GPU Codename | Feynman | Rubin (Ultra) | Rubin | Blackwell (Ultra) | Blackwell | Hopper | Ampere | Volta | Pascal |
|---|---|---|---|---|---|---|---|---|---|
| GPU Family | RF200? | VR300 | VR200 | GB300 | GB200/GB100 | GH200/GH100 | GA100 | GV100 | GP100 |
| GPU SKU | F200? | R300 | R200 | B300 | B100/B200 | H100/H200 | A100 | V100 | P100 |
| Process Tech | Intel 14A? | TSMC N2P? | TSMC N3P | TSMC 4NP | TSMC 4NP | TSMC 5nm | TSMC 7nm | TSMC 12nm | TSMC 16nm |
| CPU | Rosa | Vera | Vera | Grace | Grace | Grace | N/A | N/A | N/A |
| Memory | HBM5? | HBM4e | HBM4 | HBM3e | HBM3e | HBM2e/HBM3/HBM3e | HBM2e | HBM2 | HBM2 |
| Launch | 2028 | 2027 | 2026 | 2025 | 2024 | 2022-2024 | 2020-2022 | 2018 | 2016 |
Now the fun part begins with 3D Die Stacking. TSMC has SoIC, and Intel has Foveros Direct3D. These technologies allow chipmakers to stack chips vertically on top of each other. The other equation is advanced packaging, and here, we have TSMC CoWoS-L and EMIB-T. AMD's MI400 series is made on CoWoS-L and stacks chips on a base layer.
EMIB-T with Foveros D3D aims to achieve the same, and it looks like NVIDIA is leaning towards EMIB, now that it has allegedly achieved higher yields. Intel EMIB-T also has benefits over CoWoS, allowing more flexibility at lower costs and enhanced interconnect capabilities.
And the process node for Feynman is likely to be one of the Angstrom era nodes, TSMC A14 or Intel 14A, both of which are now set for volume ramp in 2028. Intel has pushed 14A ahead, as the previous plans were to enter HVM by 2029, but given that customers as big as NVIDIA want to get their products out as early as possible, Intel is spending big time to ensure its fabs are ready for the AI giants.
Furthermore, the recent compute demand has crippled TSMC's wafer output, as it now faces challenges to increase supply amidst a surge in compute requirements. NVIDIA is TSMC's biggest customer, but Intel is now offering NVIDIA a deal that is too hard to ignore: extra capacity, same capabilities, and chips that are made on US soil.
The Intel deal is said to include wafer and advanced packaging supply, plus it won't be limited to just one tile, but several. EMIB-T will be the interconnect that bridges these chips together, & Foveros D3D will be the solution that brings verticality to life.
Well, that is, of course, if the deal goes as planned, but based on everything that we've heard so far, it seems like a big announcement will come one day. NVIDIA is currently deep in its Rubin platform, and we expect any announcement on this matter to come by 2027 or even 2028. But once again, in all of this, Intel Foundry will seal itself as the biggest tech player in the world.
About the author: A Software Engineer by training and a PC enthusiast by passion, Hassan Mujtaba serves as Wccftech's Senior Editor for hardware section. With years of experience in the industry, he specializes in deep-dive technical analysis of next-generation CPU and GPU architectures, motherboards, and cooling solutions. His work involves not only breaking news on upcoming technologies but also extensive hands-on reviews and benchmarking.
Follow Wccftech on Google to get more of our news coverage in your feeds.
