Intel and AMD may be edging toward an unlikely partnership, with reports indicating early-stage talks for AMD to use Intel’s foundry services. If it happens, the deal could center on Intel’s 18A and 14A process technologies and would mark a major milestone for Intel Foundry Services as it courts external customers.
Industry chatter points to growing momentum behind Intel’s foundry push, with several U.S. tech giants reportedly exploring collaboration amid a broader drive to strengthen domestic chip manufacturing. Apple and NVIDIA have been floated as potential partners, and the political tailwinds in Washington only add to the appeal for companies looking to diversify production and gain negotiating leverage on future policy matters.
Rivals they may be, but Intel and AMD have worked together before. In 2018, the Kaby Lake-G project combined Intel processors with AMD Radeon graphics on a single package, proving the two can collaborate when the product makes sense. A foundry arrangement would be a different kind of partnership, but the precedent helps.
What AMD might build at Intel is still speculative. One scenario floated in reports is shifting a slice of CPU production to Intel’s 18A node—potentially even data center parts like future EPYC generations—alongside 14A in a longer-term roadmap. Today, AMD primarily relies on TSMC, including next-gen nodes such as N2, so any move to Intel would be about dual-sourcing, risk mitigation, and capacity flexibility rather than a full-on switch. Nothing is confirmed, and timelines would depend on design porting, economics, and Intel’s ability to hit performance, power, and yield targets.
That last point is critical. Intel’s 18A process is a showcase for its next-gen manufacturing, including gate-all-around transistor technology and backside power delivery, both aimed at stronger performance-per-watt and density. If Intel executes on 18A and follows with 14A on schedule, it could ease industry reluctance to shift workloads away from TSMC, especially given ongoing geopolitical uncertainties that continue to hover over global supply chains.
A deal of this nature would carry both business and political significance:
– For AMD: added supply resilience, potential cost and capacity benefits, and leverage across multiple foundry relationships.
– For Intel: validation of IFS by a top-tier CPU designer, a stronger customer pipeline for 18A/14A, and renewed credibility as a leading-edge manufacturer.
– For the broader market: more competition at the cutting edge, healthier supply diversity, and reduced single-point-of-failure risk for advanced nodes.
There are hurdles. Porting complex designs from TSMC to Intel involves requalification, IP and EDA ecosystem alignment, verification, and time. Capacity, pricing, and long-term commitments also matter. And because these discussions are still early, strategies could shift as both companies weigh roadmaps, performance targets, and cost structures.
What to watch next:
– Intel updates on 18A yields, performance-per-watt, and customer wins, followed by progress on 14A.
– Any public confirmation from AMD about diversifying manufacturing for CPUs, GPUs, or accelerators.
– Policy developments and incentives that further encourage U.S.-based leading-edge production.
– Industry signals that more high-performance designs are being evaluated for Intel Foundry Services.
Bottom line: Intel and AMD exploring a foundry partnership underscores how quickly the semiconductor landscape is evolving. If 18A delivers and the economics work, this rivalry-turned-relationship could reshape how the most advanced chips get built—and who builds them.
