Intel's 18A process was a breakthrough for the foundry division, especially with a successful Panther Lake launch, but here's why external adoption is limited for now.
Intel's PowerVia Technology Paves The Way Towards Future Customer Commitments, But Not Right Now
Intel Foundry has been struggling with customer commitments for several years now, but the division has worked extensively on the 18A process, with foundations laid by ex-CEO Pat Gelsinger. The node has seen recent success with its integration into the Panther Lake lineup, and given how Intel's latest APU platform has turned out, experts are optimistic about Intel Foundry's progress. However, there is growing concern about whether Intel will see external volume for its 18A process, and the answer lies in the fundamentals, particularly "BSPDN".
Now, what's BSPDN, and why is it being called a barrier to adoption? Well, the answer isn't too simple, but it lies in the fact that 'change is difficult to adapt to'. BSPDN refers to Backside Power Delivery Network, a one-of-a-kind implementation by Intel for its 18A node. While I won't go into the technical details, BSPDN allows chip manufacturers to route power and ground on the back side of the node, freeing up space on the front side for faster data transmission. With BSPDN, Intel essentially changed the way power delivery was traditionally handled in processes.
TechInsights, in their report on the 18A process, argues that backside power delivery is a bet that Intel has made a little 'earlier', but it could pan out to something much larger in the long term. The primary issue with BSPDN for customers is that it would require a ground-up rework of physical design methods, as the delivery network deviates from traditional logic norms. Intel's 18A is a significant leap in process design, integrating complex implementations like PowerVia and RibbonFET into a single package, which is one of the reasons behind customer hesitation.
While BSP offers long-term benefits in power integrity and scaling efficiency, it also represents a structural departure from conventional design methodologies. Adoption requires substantial design re-architecture, limiting immediate portability for customers accustomed to frontside power delivery. In contrast, competing foundries are expected to delay BSP implementation until later in the decade, with broader industry adoption anticipated around 2027.
- TechInsights
However, given that Intel is first with PowerVia, this gives them a massive edge over the likes of TSMC, which plans to introduce a similar solution with its A16 process, which is almost two generations later. And, given that Panther Lake ramp-up is on expectations, Team Blue is essentially giving its customers an indication that BSPDN contributes to power efficiency and the compute capabilities of an architecture, which is why it is expected that the 18A-P process, a derivative of the original 18A, will see more success.
It would be interesting to see how Intel and its potential customers manage the integration of the 18A family, but it seems Intel would find it more viable to target 14A-class nodes for external adoption, since by then everyone would be thinking about newer transistors and power delivery methods.
About the author: Muhammad Zuhair is a hardware and technology reporter for Wccftech, specializing in the semiconductor industry and the complex interplay between technology, manufacturing, and geopolitics. His coverage focuses on the corporate strategies and technological roadmaps of industry giants like TSMC, NVIDIA, Samsung, and Intel. Zuhair's expertise lies in deconstructing complex topics such as fabrication nodes (e.g., 2nm process), the economic impact of policies like the CHIPS Act, and the strategic development of AI infrastructure from NVIDIA, AMD and Intel.
Follow Wccftech on Google to get more of our news coverage in your feeds.
