Samsung Didn't Out-Innovate Its Rivals. It Out-Spent Them at the Worst Possible Time.

In 1983, Samsung wanted to make memory chips and didn't know how. So it did something the legend tends to leave out: it paid Micron Technology cash to teach it, then fabbed one of Micron's own 256K DRAM designs while its own reverse-engineered version was still on the bench.³ Five years later that licensed-and-borrowed beginning would seed one of the most dominant manufacturing positions on earth. The company that started life in 1938 selling noodles and dried fish to Manchuria¹ did not climb the technology ladder by inventing the rungs. It bought the first few.

The official story is that Samsung is the world's great fast follower - it watches a market grow, then enters with a better version. That is true, and it is the comfortable half. The harder half is what made the following actually win: not cleverness, but capital deployed at exactly the moment everyone else was retreating.

The miracle that started with a wire transfer

The tidy version of Samsung's semiconductor rise goes like this: in February 1983, founder Lee Byung-chul stood in Tokyo and declared the company would become a DRAM vendor, and by sheer will and engineering brilliance it shipped a 64K chip months later. The dates are real - Samsung confirms the 64K DRAM in 1983 and the 256K in 1984.⁴ What the heroic framing skips is how the gap from zero to 64K closed in a single year. It did not close on homegrown genius. Samsung licensed the know-how from Micron for a cash payment, then manufactured a Micron 256K design outright while building its own version behind it.³ That is not a footnote. It is the entire mechanism. You cannot reverse-engineer a market you are not already inside, and the license was the ticket in.

Call this the more honest read: Samsung's edge was never being first to the idea. It was being willing to acquire the idea cheaply, manufacture it relentlessly, and then keep spending into the chip industry's brutal price collapses when rivals cut back. Memory is a commodity that craters on a cycle. The company that adds capacity during the downturn owns the upturn. Samsung treated each glut not as a reason to retreat but as a discount on market share - and that countercyclical nerve, far more than any single chip, is what compounded a licensed 64K DRAM into a manufacturing empire.

What the chip share actually says

It's tempting to describe the payoff in superlatives - you'll see claims that Samsung makes half the world's microprocessors. It doesn't. In 2023 Samsung held about 7.5% of total global semiconductor revenue.⁶ The number sounds modest until you understand the shape behind it: Samsung does not spread thinly across all chips. It concentrates ferociously in memory - DRAM and NAND flash - where the fast-follower-with-capital playbook works best, because memory is exactly the commodity market that rewards the deepest pockets across the cycle. The 7.5% headline is the average of a company that is unremarkable in most chip categories and overwhelming in the few it chose to dominate. That selectivity is the strategy, not an accident of it.

The same pattern repeats in phones. Samsung wasn't first to the smartphone; Apple was. Samsung followed, entered with a better-and-cheaper-and-everywhere version, and by 2012 was the world's largest vendor. It paid for that ascent partly in court - the famous Apple verdict that headlines still report as a 'billion-dollar fine' was whittled down over six years to a final $539 million before the two sides settled in 2018.⁵ The penalty, like the chip share, is smaller and stranger than the legend. Following Apple cost Samsung real money. It also handed it a decade at the top of the market.

Isn't 'fast follower' just a polite word for copying?

The fair objection cuts the other way: that calling Samsung an imitator is too cynical, that licensing a starting point is ordinary industrial practice, and that the company has since become a genuine inventor - the firm that bends glass into folding phones doesn't owe Micron anything anymore. All true. Samsung crossed from follower to frontier years ago, and the engineering is real. But the point isn't that Samsung cheated; it's that the strategy was reproducible, and a reproducible strategy invites successors. The honest counter is the dangerous one for Samsung: the exact playbook it ran - license or reverse-engineer the technology, then out-invest the incumbent across the cycle - is now being run against it. Birkinshaw himself notes that Chinese rivals are 'playing Samsung's game.'⁸ A moat built on being the fastest, best-capitalized follower lasts only until someone follows faster with cheaper capital.

And the contest has already arrived. In 2025 Samsung shipped 241.2 million smartphones for a 19.1% global share - and was edged out of the top spot by Apple's 20%, ending a reign that had held since 2012.⁷ The company is hardly fading; the same year it posted record revenue of roughly $233 billion.⁷ But the symbolism is sharp. The world's most famous fast follower is now defending against followers of its own making, in the very category it once seized by following. The student became the teacher, and the new students are gaining.

Samsung's genius was never the chip or the phone. It was the recognition that you don't have to invent a market to own it - you have to buy your way to the starting line, then spend harder than anyone when the market is on sale. That works brilliantly, right up until someone with the same insight and a steeper cost advantage shows up behind you. Samsung wrote the fast-follower playbook so well that it taught the next generation exactly how to win. The hardest thing about being the best at a repeatable strategy is that it can be repeated.