Intel’s 18A-P starts risk production, promising 9% better CPU performance at same power

Intel’s next-generation logic process, 18A-P, has officially entered risk production, and the promise is unusually crisp: Intel claims chips made on 18A-P can deliver “9% fully routed block level performance at iso-power,” or “18% lower power at iso-performance.” This is not a vague future dream. It is happening now, in a limited production stage that comes before mass production, and Intel’s framing is that 18A-P is “the first performance enhancement in the Intel 18A family.”

The headline number matters because modern CPUs do not sell purely on raw speed anymore. They sell on a tight bundle: performance per watt, thermals, and the ability to fit into thin laptop designs and increasingly constrained handheld and client devices. If Intel’s 18A-P claims hold under real-world conditions, it is effectively a “same power, more CPU” upgrade path (or an “same performance, less power” option). And Intel emphasizes backwards compatibility to Intel 18A, meaning it is positioning 18A-P as a process improvement that does not require new fabs, new machinery, or even new chip designs. That is the kind of operational story investors and customers both like, because it reduces the churn risk that often accompanies manufacturing transitions.

So what exactly is 18A-P, according to Intel’s own slide highlights as shared by Hardwareluxx? Intel says it includes a new logic Vt option, “ULVTLL (ultra low voltage low leakage) between LVT and ULVT.” In plainer terms, Intel is adding a more granular voltage and leakage operating point, which can help designers tune chips to hit the desired power targets without giving away too much performance. Intel also points to manufacturing and tooling changes: thinner handling wafers, new materials, and an optimized EDA tool flow. The stated goal is improved thermal conductivity.

Intel also calls out changes tied to electrical reliability and performance where the wiring and power delivery really matter. The slides mention “20-40% improved thermal resistance” and “10-30% improved Via resistance at perf critical layers.” Via resistance is one of those unglamorous manufacturing realities that can quietly steal performance, especially at the layers that matter most for speed. Intel frames 18A-P as reducing that impact, which is part of how it expects the 9% performance uplift at the same power budget. In the same package, Intel ties the improvements to “backwards compatible to Intel 18A,” meaning the upgrade is aimed at slotting into the existing ecosystem rather than forcing a full re-architecture.

Timing is the other half of the story. Intel says the process is currently in risk production, which is the limited production stage that occurs before mass production begins in earnest. The practical implication for anyone planning product roadmaps is that there should be “at least a few more months” before mass production starts in earnest. That matters for procurement cycles, qualification timelines, and supply commitments. It also matters for competitive positioning because rival process transitions are often judged not only by the peak spec, but by how quickly that spec can be turned into shippable silicon that survives real thermals in real devices.

Intel’s 18A process is already in use for Panther Lake chips for laptops, and the source notes that those chips have been “pretty impressive,” which is why the industry has eyes on downstream products. The same logic extends to Intel’s Arc G3-series chips for handhelds, described as Panther Lake-based. The source adds an important nuance: it is “unlikely 18A-P will be used for these mobile chips” given they are already in full swing. That does not kill the story. It just pushes the value of 18A-P into the next product wave. The source suggests a couple of plausible paths without pretending certainty: a Panther Lake refresh, where the same chips get a performance bump, or Intel’s upcoming Nova Lake desktop chips, which would be a more natural fit for a more power-efficient process because desktop CPUs are often power and cooling limited.

There is also a strategic chessboard angle. The source mentions “a chance it could all be timed well for that upcoming Intel-Nvidia collab.” That collab angle is not confirmed as the reason for 18A-P’s timing, but it signals how process improvements can become coordination points across an ecosystem. If 18A-P reaches mass production smoothly and the performance-per-watt story is real, it strengthens Intel’s ability to offer silicon that makes platform partnerships easier, whether those platforms are gaming-adjacent handhelds or other compute-heavy designs where power budgets are the real boss.

For executives and board members, the takeaway is simple but not small: 18A-P is not just another roadmap slide. It is a process enhancement with quantified targets, and it is already in the manufacturing pipeline at risk production. If Intel can transition from risk production to mass production on schedule, maintain backwards compatibility, and deliver performance or power gains that system integrators can actually use, it keeps Intel’s silicon advantage from stalling. In a market where thermals and efficiency are increasingly the deciding factor, that is the kind of improvement that can quietly reshape product competitiveness across the next laptop and desktop cycles.