Intel has officially revealed its next-generation ´turbo cell´ standard-cell libraries as part of the planned 14A process node, targeting production in 2027. Announced at the Foundry Direct Connect symposium in San Jose, the new 14A node underscores Intel´s accelerated roadmap of delivering five advanced nodes within just four years. Currently, the company is ramping its 18A node at the Arizona facility, with 18A expected to reach full-volume manufacturing later this year and help internalize more chiplet assembly for products like Lunar Lake.
The upcoming 14A process leverages a combination of second-generation RibbonFET gate-all-around transistors and Intel´s PowerDirect, an improved power delivery network. Using cutting-edge High-NA EUV lithography, Intel estimates that 14A will offer a 15 to 20 percent performance-per-watt uplift compared to 18A. The standout feature, turbo cells, provides chip designers unprecedented flexibility by allowing both high-performance and energy-efficient cells within a single logic block. This enables precise tuning of chip characteristics to optimize speed, power consumption, and die area—capabilities especially vital in diverse application domains, from high-throughput computing to mobile devices.
Turbo cells are engineered to elevate peak CPU clock speeds and expedite critical GPU pathways, all while mitigating additional energy costs. Intel has already provided its 14A process design kit (PDK) to prospective partners, with several already plotting initial test-chip tape-outs. Supporting its technology roadmap, Intel will also deploy sophisticated chip packaging solutions like Foveros 3D stacking and EMIB, including the new high-bandwidth EMIB-T, allowing chips utilizing both 14A and 18A dies to be integrated into cohesive hybrid packages. These advances reinforce Intel´s ambitions to reclaim process leadership and deliver state-of-the-art solutions for emerging workloads in Artificial Intelligence, cloud, and high-performance computing.
