Intel spent much of the semiconductor industry’s history as the benchmark for process technology, but in recent years it ceded that leadership to rivals. The company’s response is an ambitious roadmap of new process nodes, and among them the node known as 18A has become emblematic of its bid to return to the front rank and to establish itself as a serious contract foundry.

The technologies behind the node

The push to advanced nodes such as 18A brings together two important innovations. The first is a new transistor architecture often described as gate-all-around, in which the channel is surrounded on all sides by the gate for better control of current at very small dimensions. This succeeds the fin-based transistors that dominated previous generations.

The second is a change in how power is delivered to the chip. Traditionally, both signal and power wiring sit above the transistors. Backside power delivery moves the power network to the underside of the wafer, freeing space for signal routing and reducing losses. Together, these approaches aim to improve performance and efficiency at the leading edge.

Foundry ambitions

Beyond making its own products, Intel intends to manufacture chips for external customers, competing directly with established foundries. This is a substantial shift in business model and requires:

• Design tools and support tailored to outside customers
• Predictable, well-characterised process technology
• Trust that competitors’ designs will be handled securely
• Yields and costs competitive with incumbents

A demanding turnaround

Regaining process leadership is difficult, and success will be judged by delivered products and customer adoption rather than roadmaps alone. Even so, a credible third leading-edge foundry would meaningfully change the industry’s competitive landscape.

For VLSI engineers, Intel’s 18A effort highlights how transistor architecture and power delivery are evolving, knowledge that grows more valuable as these techniques spread across the industry.