Intel earnings release had good news for tool makers.
Intel slows its most ambitious node—and signals where the real battle is for now
Jingyue Hsiao, DIGITIMES Asia, Taipei
Friday 23 January 2026
Intel is sharpening its manufacturing strategy around yield improvement, selective capital spending, and differentiated advanced packaging, while keeping major 14A capacity investments on hold until customer demand is secured.
Capex shifts from space to tools as supply tightens
Intel's capex strategy is becoming more targeted rather than expansive. CFO David Zinsner pushed back against the idea that Capex is simply "flat to slightly down," saying the picture is more nuanced. The company is materially reducing spending in some areas, particularly physical space, while reallocating capital toward manufacturing tools.
This shift reflects Intel's immediate bottleneck: tool availability rather than clean-room footprint. Zinsner said Intel plans a notable increase in tool spending in 2026 compared with 2025 to address supply shortfalls, underscoring that future capex growth will be directed at enabling output rather than building new shells. The approach signals a focus on faster time-to-capacity and tighter return discipline after years of heavy fab construction.
Advanced nodes ramp selectively, with 14A on hold
Intel management emphasized disciplined execution across advanced nodes rather than blanket capacity expansion. Zinsner said the company is "aggressively getting tools on Intel 7, Intel 3, 18A" and increasing wafer starts as quickly as possible where demand and readiness align.
In contrast, Intel is deliberately holding back on 14A capacity. Zinsner said the node is "really linked to foundry customers," and it does not make sense to build significant capacity until customer demand is secured. For now, spending on 14A is limited largely to research and development.
Management expects clearer customer decisions in the second half of this year and into the first half of next year. Only then, Zinsner said, will Intel "start to unlock the spend on 14A," reinforcing a demand-driven approach to foundry expansion rather than speculative buildout.
High-NA EUV positioned squarely for 14AIntel also clarified the role of high-numerical-aperture EUV in its roadmap. When asked whether recent demonstrations were aimed at a future 10A node, Zinsner said high-NA lithography "will be part of our 14A process," with different variants of 14A incorporating the technology.
This signals a concrete commitment to deploy advanced patterning at 14A, rather than deferring it further out. The implication is that Intel expects high-NA to be a key enabler for performance, density, and manufacturability as 14A moves toward higher-volume production, strengthening its competitiveness in leading-edge foundry services.
Yield improving steadily, but not yet best-in-class
CEO Lip-Bu Tan struck a cautiously optimistic tone on manufacturing execution. He said Intel is seeing yield improvements of roughly 7% to 8% per month, while continuing to focus on reducing variation and defect density to deliver consistent wafers.
However, Tan was explicit that Intel has not yet reached industry-leading standards. "I see improvement, but still not quite to that industry-leading standard yet," he said, underscoring that execution risk remains even as progress is evident. The comment reflects both operational momentum and the reality that sustained yield improvement is critical before Intel can fully close the gap with leading competitors.
Advanced packaging emerges as an early revenue lever
Advanced packaging is emerging as one of Intel's clearest near-term differentiators. Zinsner said management previously expected packaging opportunities to be measured in the hundreds of millions of dollars. Still, early customer engagements now point to revenue "well north of US$1 billion" for many of these programs.
He described Intel's packaging technologies as highly differentiated and particularly well-suited for AI workloads. Tan echoed this view, calling advanced packaging—especially the T variant—a "very big differentiator." He added that some customers are willing to prepay due to severe supply shortages, a sign of both high demand intensity and strategic partnership alignment.
Notably, management expects advanced packaging revenue to materialize before large-scale wafer revenue, positioning it as an early validation of Intel Foundry's value proposition, even as leading-edge wafer commitments are still being finalized.
Supply shortages complicate manufacturing and allocation
Tan said first-quarter weakness across both client and data center segments is driven by supply, not demand. Intel is prioritizing wafer supply for data center customers to meet strong AI-driven demand but cannot fully abandon the client market, forcing difficult tradeoffs that constrain shipments across both businesses.
He described the first quarter as a supply trough, noting that inventory buffers built in late 2025 have largely been depleted. Intel is now operating "hand to mouth," shipping only what it can produce and deliver immediately. Management expects improvement in the second quarter but warned that constraints will linger through the year.
Falling margins remain the near-term reality
Against this backdrop, gross margin pressure remains unavoidable. Tan acknowledged that the guided 34.5% gross margin level is "by no means an acceptable" outcome. Intel's first goal is to return to 40%, with higher targets to follow once that threshold is achieved.
Zinsner cautioned that near-term incremental margins are weak, as mix, supply timing, and the early cost curve of new products weigh on results. While management expects margins to recover as yields improve and throughput increases, the message was clear: falling margins are the price Intel is paying in the near term to rebuild its manufacturing base, scale 18A, prepare 14A, and position advanced packaging as a growth engine for its foundry ambitions.
Article edited by Joseph Chen
digitimes.com |
