(Credit: Annabelle Chih/Bloomberg via Getty Images)
Intel's CEO this weekend reiterated plans to launch "new products" with Nvidia.
Lip-Bu Tan awarded Nvidia CEO Jensen Huang with an Honorary Doctorate in Science and Technology at Carnegie Mellon University on Sunday, "for his outstanding contributions to accelerated computing and Artificial Intelligence," Intel's CEO tweeted, adding that Intel and Nvidia are "collaborating to develop exciting new products."
It comes several months after Intel and Nvidia teamed up to put RTX graphics directly into Intel PC chips; Nvidia also invested $5 billion in Intel via a stock buy.
That followed a rocky year for Intel that saw the US government taking a 10% stake in Intel by using $8.9 billion in federal subsidies following a run-in with President Trump and mass layoffs.
As TechPowerUp notes, Intel and Nvidia are expected to developer integrated third-party graphics for x86 SOCs. Codenamed "Serpent Lake," they're designed to compete with AMD's Strix Halo line of APUs, which combine high-end CPU cores with powerful onboard graphics to challenge entry-level dedicated GPUs on mobile and handheld gaming systems.
WCCFTech speculates that it will be an offshoot of Intel's next-generation mobile lineup, Titan Lake, but we might not see it until 2028. It will include a custom Nvidia GPU, potentially based on the Vera Rubin architecture or even a generation beyond that.
Another Nvidia and Intel collaboration is likely to include custom Intel Xeon processors alongside Nvidia Grace CPUs and Vera CPUs, in order to operate as part of its HGX AI server nodes. Intel's upcoming Diamond Rapids Xeon CPUs will integrate Nvidia's NVLink technology, too, TechPowerUp reports.
The real potential of this deal isn't in any single product line, but in fabrication. Intel has been ramping up its fabrication business in America and looking to sell capacity for its new 18A and next-generation 14A nodes. Nvidia has been seeking additional partners for GPU fabrication to reduce its reliance on TSMC. It's already set to use Intel's advanced packaging for its next-generation Feynman AI accelerators, and is reportedly considering using it to fabricate the I/O die for the chips.
None of this is happening in the short term, but that doesn't stop these announcements and re-commitments from having an impact. Investors are looking at Intel as one of the next hot vehicles as CPU demand becomes more apparent in the AI supply chain, and its stock has risen dramatically in recent weeks. Intel and Nvidia reconfirming their partnership, even if no hardware is just around the corner, is one way to keep that hype rolling.
It's a familiar tactic. Tan's X feed is mostly him touting partnerships with the likes of Google, Cisco, and even Elon Musk regarding his Terafab plans.