How Semiconductor Shortages Changed the Tech Industry

The great chip shortage of 2020–2022 didn’t just delay your new car or PlayStation. It rewired how the tech industry thinks about supply chains, capacity, and where the real bottlenecks are. For decades, the assumption was that chips would get cheaper, smaller, and more abundant. Just-in-time inventory and globalised production were the norm. Then a pandemic, a drought in Taiwan, and a fire at a key fab turned “we’ll get more next quarter” into “we might get some in 2024.” The industry is still adjusting.

What Broke

Semiconductors are the foundation of almost everything that runs on electricity: phones, cars, servers, medical kit, and industrial controls. The supply chain is extraordinarily concentrated. A handful of companies make the most advanced logic chips; a few regions—Taiwan, South Korea, and increasingly the US and Europe—host the fabs. When demand spiked (work-from-home, cloud, 5G, EVs) and supply was hit by lockdowns, accidents, and underinvestment in mature nodes, the whole system jammed. Lead times stretched from weeks to a year or more. Automakers idled lines. Tech companies redesigned products to use whatever chips were available. Everyone learned the hard way that “supply chain” wasn’t a back-office concern—it was the constraint that could cap growth overnight.

From Just-in-Time to Just-in-Case

Before the shortage, the dominant logic was lean inventory. Why tie up capital in stock when you could order what you need, when you need it? The shortage exposed the downside: when the pipeline breaks, you have nothing. Companies that had diversified suppliers or held buffer stock fared better. Those that had single-sourced critical components or assumed infinite elasticity got burned. The result has been a shift toward “just-in-case” thinking—carrying more inventory, qualifying second sources, and sometimes paying a premium for guaranteed allocation. That costs money and complicates operations, but the cost of a line-down event has been repriced. Risk is no longer something you optimise away to zero; it’s something you hedge.

Reshoring and Geopolitics

The shortage also accelerated a trend that was already in motion: bringing chip production closer to home. The CHIPS Act in the US and similar initiatives in Europe and elsewhere are explicitly about reducing dependence on a single geography. The motivation isn’t only economic—it’s strategic. Semiconductors are seen as critical infrastructure. If most of the world’s advanced logic is made in one place, that place becomes a single point of failure and a lever in geopolitical disputes. So we’re seeing billions in subsidies for new fabs in Arizona, Ohio, Germany, and Japan. Whether that translates into a more resilient supply chain or just duplicate concentration remains to be seen. Fabs take years to build and require deep pools of talent. Geography alone doesn’t solve the capacity problem.

Who Gets the Wafers

When capacity is tight, allocation matters. The big players—Apple, Qualcomm, AMD, NVIDIA—lock in long-term deals and often pay upfront for capacity. Smaller customers and those in less glamorous segments (automotive, industrial) found themselves at the back of the queue. Car chips are often made on older, “mature” nodes that were supposed to be cheap and plentiful. When fabs prioritised high-margin logic and memory, mature-node capacity didn’t keep up. The auto industry learned that it couldn’t assume it would always get what it needed. The response has been more direct engagement with fabs and, in some cases, commitments to take capacity for years. The tech industry has always had tiers of customers; the shortage made that hierarchy visible and painful for those at the bottom.

Design and Substitution

Another lasting change is design flexibility. Companies that could redesign a product to use a different chip, or a different package, had more options when their first choice was unavailable. That has pushed more teams to think in terms of substitutability from the start—second sources, pin-compatible parts, and modular designs that can absorb supply shocks. It’s not free: it adds design time and sometimes performance or cost trade-offs. But the shortage proved that flexibility is an asset. The same goes for software: when hardware is constrained, the ability to do more with less—better compression, smarter scheduling, fewer redundant components—becomes a competitive advantage.

The New Normal

Shortages have eased. Lead times have come down. But the industry hasn’t reverted to the old playbook. Inventory levels are higher. Supplier relationships are more strategic. Geopolitical risk is part of every capacity decision. And the next shortage—whether from another pandemic, a natural disaster, or a conflict—is a matter of when, not if. The lesson of the last few years isn’t that we fixed the problem. It’s that we learned how fragile the system was and started to take resilience seriously. The tech industry runs on silicon. It now knows that the pipeline that delivers that silicon is as important as the design of the chip itself.