Q&A with Intel EVP Dr. Ann Kelleher on lessons learned from the company's misadventures with 10nm process nodes, its vision for Foundry Services, and more
This AnandTech Q&A closes that loop from the inside: Kelleher explains what went wrong at 10nm and lays out the vision for Foundry Services — the merchant-foundry ambition that Pat Gelsinger would later detail in his own Q&A on Intel Foundry, IDM 2.0, 18A, and reluctant customers. It matters because Intel is asking external customers to trust a company still narrating its own execution failure.
First-order effects
Kelleher puts an official post-mortem on the 10nm era, giving customers and partners Intel's own account of the execution failure rather than press speculation.
Foundry Services gets a named executive champion, signaling Intel intends to run the foundry business as a distinct offering rather than an internal capacity plan.
Second-order effects
Prospective foundry customers now have a benchmark for judging whether Intel's roadmap commitments are credible against TSMC and Samsung, making Intel's node-execution track record a sales variable, not just an engineering one.
Rival foundries face a competitor willing to discuss its failures openly — a transparency posture that raises expectations for how all leading-edge suppliers communicate slippage.
Third-order effects
If the pattern holds, Intel's identity shifts from integrated device maker to a two-sided organization — product company plus merchant foundry — with node credibility as the hinge between them.
The long arc from 10nm recovery to Foundry Services also foreshadows a generational handoff: by 2025 Kelleher had moved to strategic adviser ahead of retirement as Intel reorganized Foundry leadership, meaning the lessons she codifies here become institutional knowledge rather than personal ones.
The trend: Leading-edge chipmaking is consolidating around a handful of trusted foundries, and Intel's decade-long effort to convert its 10nm failure into foundry credibility is one data point in that consolidation.
AMD Sr. fellow Rohit Verma, Lead SoC Architect for Radeon Discrete GPUs, has been poached to be Lead Product Architect for Intel Accelerated Computing Group “The brain drain, well that's changed. We now have brains coming back” Pat Gelsinger @PGelsinger @Rajaontheedge $INTC $AMD …
Intel's ‘absurdly bullish’ long-term forecast sends stock to lowest point https://a.msn.com/... // Now is a time when analysts and the company diverge because the timelines are too long for analysts and too short for the heroes building at Intel (no stock advice follows) 1/
5/ Having been close to these choices and made our own bets, I definitely can empathize on two levels. First, they do need time. New fabs don't just pop up in time for a shareholder meeting/analyst day. These are huge, expensive, and take time.
13/ That's no guarantee. The only thing for sure is saying “absurdly bullish” is just a mismatch of skillsets/expertise and mindset. Right now everyone is bearish in markets. Product people are naturally bullish. This is just going to play out. 🙏 // END
10/ I've certainly been on the side of describing how intel choices have got them to where they are. They were disrupted by ARM. But the way chips work, that hardly forecloses the ability to leapfrog. In fact the business of chips is literally a business of leapfrogging.
Many Intel stories in Taiwan: Intel's roadmap indicates it will beat TSMC to 2-nanometers in 1st half of 2024, vs TSMC in 2025. Pundits plan to scrutinize Intel's 4nm, sub-4nm yields, and say Intel is no threat to TSMC over the next 2-3 years. $INTC $TSM https://www.chinatimes.co…
The Intel skeptics are nitpicking. The only thing that matters is Intel will leapfrog TSMC to RibbonFET/PowerVia at 20A and leapfrog TSMC to high-NA EUV at 18A. Intel will be on second-gen RibbonFET before TSMC's first gen. Major IFS selling point. https://twitter.com/...
With this interview, I'm completing my final day here at @AnandTech. It's been a great 10 years working with the team, with @ryansmithAT, @anandshimpi, @bretthowse, @gavbon86 + all. My next venture will be in tech - hope to give more information next week. I'm not going far! 😎🤘 h…
Perfect description of Ann Kelleher: “What I loved about speaking to Ann is the element of quiet but striking determinism in the way she spoke...” Insightful Q&A with Intel's GM of Technology Development. #IAmIntel https://twitter.com/...
Dr. Ann Kelleher on High-NA EUV in I18A (1/4): > We're aiming to introduce it in more in 2025, and we're setting up our processes so that if for some reason High-NA is not ready, then we will be able to continue without it. (...) https://www.anandtech.com/...