Intel's 18A-P process enters risk production with 9% performance lift

At a glance:

• Intel says 18A-P, a backward-compatible enhancement of 18A, has entered risk production after more details at VLSI 2026.
• The process promises a 9% frequency improvement at the same power, or an 18% power reduction at the same performance, at 0.75V.
• New W1, W1.5, and W3P transistor options, plus the ULVTLL threshold-voltage pair, expand design choices for 18A chips.

What changed at Intel's 18A node

Intel has added more detail to its 18A-P process after a paper published earlier this year, with the latest update coming at VLSI 2026. The performance-optimized node is an enhancement of 18A, the same advanced process family Intel is using in products such as Panther Lake and Xeon 6+. Intel's latest figures put the upside at a 9% improvement in performance at the same power, or an 18% reduction in power consumption at the same performance level. The company also confirmed that 18A-P has entered risk production, a meaningful step beyond research disclosure. Risk production sits just before high-volume mass production. It is a low-volume manufacturing stage where Intel can produce full wafers of 18A-P on a standard production line, but with a limited scope designed to gather data on defect rate, performance, and variability before full production begins. On advanced logic nodes, risk production usually leads mass production by 12 to 24 months, though Intel is not starting from an entirely new node here, so the timeline could be tighter. That matters because 18A is already ramping in two U.S. fabs, and Intel says defect rates continue to drop even as the company has faced criticism over poor 18A yields.

The drop-in compatibility angle

18A-P is a revision of 18A rather than a clean break from it. Although it carries new transistor designs, those designs live in the same libraries at cell heights of 180mm for High Performance and 160mm for High Density. The key manufacturing and design promise is backward compatibility with 18A designs. Intel says designers can port to 18A-P without making changes, and while some of the new transistor options could spur a design change, that is not required. For chip teams already working on 18A, that compatibility reduces the risk of treating 18A-P as a separate process with a fresh validation burden. Anything built on 18A can be built on 18A-P with minor performance benefits, while teams that want more upside can explore the new transistor and threshold-voltage options.

Performance, power, and new transistor options

Intel arrived at its headline numbers by testing on a standard Arm core subblock. The 9% frequency increase and 18% power reduction are specified at 0.75 volts, and Intel says 18A-P retains a frequency and power improvement even as voltage moves outside that 0.75V mark. The process update adds three transistor designs to Intel's library. W1 is available in the 180mm cell-height library, where it was previously available in the 160mm library, while W1.5 is available in the 160mm library, and the enhanced W3P design is available in both libraries. W1 and W1.5 are narrow designs optimized for low-power usage, filling gaps in power-optimized parts of Intel's library. W3P is the standout addition: it is a new dual-contact transistor with Power Boost, Intel's name for the technique. 18A already uses backside power delivery through PowerVia, which routes power on the back of the wafer to free front-side signal wiring and reduce thermal resistance. W3P adds contacts on both the front side and backside, reducing parasitic resistance and enabling higher drive current to speed up switching. Even the standard W2 and W3 transistors see a boost in ring oscillator frequency in a ring oscillator, which moves an electrical signal through a ring of inverters, when moving from 18A to 18A-P. The biggest frequency improvement comes from W3P, while W1 pushes 18A-P down to lower capacitance levels for energy-optimized designs.

More design flexibility through VT choices

Intel is also adding a new VT, or threshold-voltage, pair to its lineup. The usual set includes four flavors: HVT, SVT, LVT, and ULVT, standing for high, standard, low, and ultra-low threshold voltage, respectively. Threshold voltage is a core tradeoff in chip design. The lower the threshold voltage, the less power a transistor needs to activate, but the more power it leaks; ULVT transistors are the most performant but leak the most power, while HVT transistors are the least performant but leak the least. That is why Intel's new ULVTLL option matters. ULVTLL, or Ultra-Low Voltage Threshold Low Leakage, lives between ULVT and LVT, offering better performance than LVT but lower leakage than ULVT. Like the new transistor options, ULVTLL gives designers more flexibility when designing a chip for 18A-P. The same update also brings a 20% to 40% improvement in thermal resistance and a 10% to 30% improvement in via resistance at 'perf critical layers'. Intel says the thermal-resistance improvement comes from grinding the wafer down with advanced EDA tools for better thermal conductivity. For performance-critical layers, lower via resistance can help preserve the speed and efficiency gains that the new transistor options are meant to deliver.

Why risk production matters now

The practical question is whether 18A-P can turn Intel's process improvements into production silicon without disrupting the 18A roadmap. Intel already uses 18A in Panther Lake and Xeon 6+, and it is reportedly in talks with Apple and Nvidia to build on 18A. For internal product teams, 18A-P offers a potentially lower-risk route to better frequency, lower power, and improved thermal behavior without forcing a full redesign. For external foundry customers, the same compatibility story could make Intel's process more attractive if risk-production data confirms lower defect rates and predictable variability. The watch items are straightforward: whether Intel can keep defect rates moving in the right direction, whether the W3P and ULVTLL options deliver usable gains in real designs, and whether 18A-P can move from full-wafer risk runs to broader manufacturing quickly enough to support the next wave of 18A products.