Intel Foundry advances Intel 18A-P toward production as it outlines longer-term process R&D at VLSI

Intel Foundry used the 2026 VLSI Symposium in Santa Clara, California to update partners on the status of its leading-edge roadmap and to lay out how its longer-term silicon research is meant to translate into performance and efficiency gains. The company’s central message was progress on Intel 18A-P, the first performance enhancement in the Intel 18A process family, which Intel said has entered “risk production” while meeting the timeline it previously shared with customers and partners.

Intel 18A is positioned by the company as Intel Foundry’s leading-edge process node, with a production timeline it describes as on track for 2025. In its VLSI update, Intel linked the move toward production readiness to specific technology pillars: RibbonFET, a transistor design intended to improve switching and drive capability, and PowerVia, a power-delivery approach designed to support higher performance and better energy efficiency as chip complexity increases.

Intel also said it brought gate-all-around (GAA) transistors and backside power delivery (BSPD) to market last year with Intel 18A. At VLSI, Intel engineering teams discussed how those features are expected to provide a foundation for improved performance, energy efficiency, and scaling for future logic designs, effectively tying the current node to the needs of next-generation compute devices.

In a quote included in the release, Naga Chandrasekaran, executive vice president and general manager of Intel Foundry, described the VLSI updates as a announcement of long-term commitment to leading-edge process innovation. He called the effort “a journey,” adding that Intel appreciates the opportunity to share progress on Intel 18A-P and its longer-range research and development work, while acknowledging that more engineering work remains.

According to Intel, Intel Foundry enables Intel 18A-P performance, power, and design benefits through a combination of transistor, interconnect, and design-technology co-optimizations. In plain terms, the company is describing a tightly coupled process where changes in the underlying transistor and wiring, along with how designers use those features, are intended to work together rather than in isolation.

The release framed the VLSI discussions around both near-term milestones and longer-horizon R&D. Intel said it presented long-term research updates across several areas important to future silicon scaling, though it did not provide additional numeric targets or detailed technical outcomes in the text made available for this story. The company instead emphasized that it is investing in process innovation designed to keep improving the efficiency and scalability of logic chips as demand grows around AI computing.

Beyond the process node itself, Intel’s messaging at VLSI reflected the broader foundry challenge of aligning new manufacturing steps with customer adoption. Risk production typically refers to an early manufacturing phase in which a process is tested with the expectation that it may still undergo refinements, reducing uncertainty before full-scale production. By saying Intel 18A-P is in risk production and is aligned with prior timing, Intel is effectively communicating that the company believes it is close enough for customers and partners to plan designs around the next performance step in the Intel 18A family.

Still, Intel’s release did not disclose performance, power, or yield metrics in the excerpt available here, nor did it specify which external customers have begun design transitions for Intel 18A-P. It also did not enumerate the exact long-term R&D areas in the portion of the document provided, beyond referencing “several areas important to future silicon scaling.” As with many leading-edge process updates, investors and industry observers will likely look for follow-on details such as confirmed ramp timing, yield improvements, and customer tape-out or product integration milestones.

What to watch next is whether Intel provides additional specificity as Intel 18A-P moves from risk production toward full production, and whether the company later quantifies the impact of RibbonFET, PowerVia, GAA, and BSPD on real workloads. In the near term, attention is likely to focus on foundry adoption indicates, including how quickly ecosystem partners can convert designs to the updated process and how that translates into measurable efficiency and performance improvements for AI-focused compute systems.