The Courage to Cut: Intel and the Discipline of the Foundry
They didn’t lose because they stopped trying. They lost because they tried to do everything. A company can survive a bad quarter. It can survive a bad...
They didn’t lose because they stopped trying. They lost because they tried to do everything. A company can survive a bad quarter. It can survive a bad product. What it can’t survive is a blurred identity. Intel’s identity used to be simple. Make the best chips. Make them in your own factories. Ship the future on your own schedule. Then the ground shifted. Manufacturing got brutal. The physics got stubborn. The timelines slipped. Competitors didn’t just catch up. They rewrote the map. And Intel did what large, proud companies often do when they feel the world tightening around their ribs. They added. More initiatives. More side quests. More “strategic options.” But in the end, the existential threat wasn’t a rival. It was dilution. A great engine starving because it kept feeding everything except the flywheel.
When the house is on fire, you don’t redecorate. You choose what to save.
The constraint: you can’t buy your way out of physics
Semiconductors don’t reward wishful thinking. They punish it. They punish it with yield. With delays. With missed windows that never reopen. Intel’s core constraint wasn’t “market share.” It wasn’t “branding.” It wasn’t even “talent.” It was time—measured in process nodes, product cycles, and credibility. In chips, trust is a currency. Once you spend it, you don’t get it back with a press release. You get it back with wafers that work. So Intel faced a stark question: Do we exist to be a portfolio… …or to be a factory-led product company again? That question sounds abstract until you translate it into reality: A factory is not an app. You can’t iterate a fab. You commit. You pour concrete. You wait. You bleed cash for years before you see a single good die roll off the line. And if you want to lead, you need more than a fab. You need a system. A strategy that can stare down the next decade and not flinch.
The bold move: sell the distraction. feed the furnace.
In October 2020, Intel agreed to sell its NAND memory and storage business to SK hynix for 20 billion for two new fabs in Arizona, explicitly tied to the IDM 2.0 evolution and future foundry capacity. (d1io3yog0oux5.cloudfront.net) This wasn’t a “cost optimization.” This was a furnace build. A decision to compete where the barriers are real—and where the rewards are permanent.
The sacrifice: saying no while the room begs for yes
Divesting NAND wasn’t just about money. It was about permission. Permission to stop pretending that everything fits. Permission to say: we can’t win every war. We will choose our war. This is what focus looks like when it hurts. You sell the thing you built. You hand customers to someone else. You let go of employees, history, pride. You accept the headlines. You accept the second-guessing. Because the alternative is worse. The alternative is a company that keeps its options open until it has no options left. Constraint forces clarity. And clarity forces sacrifice. That’s the chain. Intel’s sacrifice carried an additional sting: the NAND business wasn’t merely a “random asset.” It was storage. It was a story Intel could tell end-to-end. It was a proof point that Intel could be more than CPUs. But “more” is not a strategy. “More” is how you die tired.
The most dangerous phrase in a successful company is: “We can do that too.”
The execution: make manufacturing a product again
IDM 2.0 wasn’t nostalgia. It was architecture. It acknowledged something modern and uncomfortable: No one ships alone anymore. The best products now mix. Different tiles. Different process technologies. Different suppliers. Different timelines. The old purity—design everything, make everything, control everything—became brittle. So Intel embraced a more flexible model while still anchoring on its signature strength: manufacturing. Build internally. Use outsiders when they’re better for a given job. But keep the identity intact: Intel fights on the factory floor. Then came the part that separates “strategy” from “courage”: They didn’t just modernize their own pipeline. They chose to serve others. Intel began building the scaffolding around IFS—because a foundry is not just a cleanroom. A foundry is a relationship. A promise. A support model. A developer ecosystem in hard hats. In 2022, Intel launched a $1 billion fund to help build a foundry innovation ecosystem—explicitly admitting a truth that proud companies struggle to say out loud: Not all good ideas start inside your walls. (intc.com) That sentence matters. It’s humility turned into operating system. It’s the difference between “we will catch up” and “we will rebuild trust.”
Why this is timeless (and why it’s not about chips)
Most industries don’t have fabs. But every industry has a furnace. A core engine that makes the company itself—not just its revenue. The timeless lesson in Intel’s move isn’t “sell NAND.” It’s this: When the constraint becomes existential, you stop managing the perimeter and you reinforce the center. You don’t ask, “What else can we do?” You ask, “What must we stop doing so the one true thing can breathe?” In retail, that furnace might be distribution discipline. In airlines, it might be fleet simplicity and operational reliability. In healthcare, it might be outcomes and trust. In media, it might be the integrity of the editorial voice. In software, it might be product coherence. In manufacturing, it might be process excellence. Whatever it is, your furnace cannot run on leftovers. It requires fuel. Capital. Talent. And above all, attention. Divestitures look like finance. But the best ones are design. They are subtraction in service of meaning.
The quiet strength: choosing a long road on purpose
The hardest strategies aren’t bold because they’re flashy. They’re bold because they’re slow. Because they refuse the dopamine of quick wins. Intel’s path demanded patience—new fabs don’t care about quarterly narratives. They care about earthmoving, supply chains, permits, tooling, yields, and time. And time is indifferent to ambition. Yet that is precisely why the strategy is iconic. It treats the future as something you build, not something you announce. And it treats focus as a moral act. Not a PowerPoint posture.
Real focus is not “prioritization.” Real focus is grieving what you won’t become.
A simple framework you can steal
If you want a Jobsian strategy, don’t start with the market. Start with the constraint. 1) Name the constraint in one sentence. If you can’t say it simply, you don’t understand it. “Time-to-trust.” “Cost-to-serve.” “Complexity.” “Regulatory drag.” “Churn.” “Supply volatility.” 2) Protect the furnace. List the three activities that feed your core engine. Then circle the one that makes everything else possible. That is the thing you fund first. 3) Cut the elegant distractions. Not the obviously bad ideas. The good ones. The ones that make you feel modern. The ones your smartest people love. The ones that win internal debates. Cut them anyway. 4) Reinvest in the boring, expensive truth. The work that can’t be faked. The work competitors can’t copy quickly. The work that compounds. Intel did this in the language of semiconductors. But the shape of it applies everywhere.
The closing: the cleanest strategy is a vow
At some point, every company meets the moment where it must choose: Will we be interesting… …or will we be great? Greatness asks for a vow. A narrowing. A refusal. A willingness to disappoint people in the short term so you can serve them in the long term. Intel’s story—selling a major business, naming a constraint, rebuilding the center—isn’t perfect. But it is clean. And clean is rare. Because clean requires courage. And courage requires sacrifice. And sacrifice is the price of focus. In the end, the strategy is not the slide. It’s the cut. It’s the concrete. It’s the years. It’s the decision to bet the company on the thing only the company can be.