The King of South Korea's Dating Arena, How SK Hynix Overtook Samsung?

Original Author: TechFlow 

At Seoul's matchmaking agencies, a peculiar trend has recently emerged.

Some men attending blind dates deliberately tuck their business cards into the innermost pocket of their suit jackets. Only after confirming that the other person is "decent enough" do they carefully pull out the card emblazoned with the company name. The card bears four English letters: SK Hynix.

Source: South Korean variety show

Kang Eun-sun, a senior executive at Korean matchmaking firm Gayeon, publicly told the media that since the start of the semiconductor super cycle, the popularity of employees from Samsung Electronics and SK Hynix has been steadily rising. "The market clearly prefers engineers with significantly higher actual incomes over lawyers whose earnings have declined." There's even a joke circulating on social media: "When Hynix employees go on blind dates, they humbly say they work at Samsung. Only when they meet someone with good character do they confess they actually work at Hynix."

What turned a work uniform into a "date-night armor" is a set of numbers that would make any worker around the world green with envy.

In 2025, SK Hynix's operating profit reached 47.2 trillion Korean Won. According to a new agreement reached with the union last September, 10% of operating profit goes into the employee bonus pool. Averaged across 35,000 employees, each person could receive about 140 million Korean Won, equivalent to roughly 650,000 RMB.

In the first quarter of this year, SK Hynix's operating profit surged over 400% year-on-year to 37.6 trillion Korean Won. Based on forecasts from analysts in different countries, this year's operating profit is expected to fall between 210 trillion and 250 trillion Korean Won. By this estimate, the average annual bonus per person this year could range from 2.9 million to 3.3 million RMB.

International investment bank Macquarie Securities further predicts that if operating profit reaches 447 trillion Korean Won in 2027, the average bonus per person could hit as high as 1.29 billion Korean Won, approximately 6.1 million RMB.

A more compelling story than the "6.1 million average bonus" is this: This company has long been the second fiddle in South Korea's semiconductor industry, the little brother standing behind Samsung.

What did it do right that made Samsung, the giant whose screens and chips even Apple relies on, fall from its position as the global storage霸主? Let's see.

2008: Recovery from the Brink of Bankruptcy

Rewind to 2008, and no one would have placed the label "future dominator" on Hynix.

Its predecessor was Hyundai Electronics. When the internet bubble burst in 2001, DRAM prices plummeted. The company was saddled with a staggering $14 billion debt, was taken over by creditors, and entered a five-year "workout program," a state akin to "restructuring trusteeship" in the Korean context. For five full years, factory operations, R&D budgets, and staffing levels were all severely constrained.

By 2007, Hynix had finally crawled out of this "trusteeship" state, but it was weak and barely surviving.

Right at that moment, across the Pacific, a company called AMD came knocking.

AMD was in a tough spot too. It was the second player in the GPU market, struggling to breathe under NVIDIA's pressure in the gaming card space. Its researcher, Bryan Black, was working on a peculiar concept called "High Bandwidth Memory" (HBM). This technology stacked multiple DRAMs vertically like building floors, connecting them through a technology called TSV (Through-Silicon Via).

Why do this? Because AMD spotted a problem that most had overlooked: CPU/GPU computing speeds were getting faster, but the data transfer speed of memory couldn't keep up. The computing units would often finish a calculation and then just wait idly for the memory to deliver the next batch of data. The industry calls this bottleneck the "Memory Wall."

To put it in a rough analogy: Imagine a super chef who can chop ingredients for 10 dishes per second, but the food runner can only bring in ingredients for 2 dishes per second. The result is that the chef spends 80% of the time waiting for ingredients. No matter how fast the computing chip is, if data can't get in, it's just spinning its wheels.

AMD's idea was straightforward: Instead of widening the data path horizontally (which is the traditional DDR approach), why not make the memory "taller"? Stack layers vertically on the chip, allowing data to travel shorter distances through a wider bus. This vertically stacked "little tower" is HBM.

The concept sounded great. The problem was that in 2008, there was no AI demand, no large model training, and no so-called "computing revolution." The only visible application for HBM at the time was for high-end gaming graphics cards. The market was tiny, the manufacturing process was extremely difficult, and the cost per die was far higher than regular DRAM.

AMD searched around. No one wanted to take it on. Samsung was unwilling; it was fully committed to another technology path, HMC (Hybrid Memory Cube), a different vertical stacking solution it was developing with Micron. Micron also declined, following Samsung's lead.

The only one willing to take the project was Hynix, the company barely recovered from bankruptcy that couldn't afford to be picky about any big order.

In 2009, Hynix officially launched HBM R&D. It wasn't until 2013 that the first HBM chip was born in Hynix's factory in Icheon, South Korea.

Back then, who could have imagined that 15 years later, this chip would become the most sought-after item by all AI giants, with production capacity "sold out until 2030"?

No one could, not even Hynix itself.

2012: A Chairman's Gamble

In 2012, another key figure entered the story.

SK Group Chairman Chey Tae-won led a consortium formed through SK Telecom to acquire a 21.05% stake in Hynix from creditors for approximately 3.4 trillion Korean Won (about $3 billion). This semiconductor company was henceforth known as SK Hynix.

What kind of person is Chey Tae-won? A passage from the Korean book *Super Momentum* describes the event: After the acquisition, he held one-on-one meetings with 100 of Hynix's top executives in a short period. His first action wasn't layoffs or cost-cutting, but integrating SK Group's management system with Hynix's technical capabilities, and then re-opening the funding for FAB (wafer fab) investments and process improvements that had been suspended.

The key was timing.

From 2012 to 2014, the entire DRAM industry was still under the shadow of the 2011-2012 memory downcycle. All rational financial models were telling management, "This is an industry trough, be conservative." But Chey Tae-won made the opposite decision: expand investment.

Even more critical was HBM.

Over the 11 years from 2011 to 2022, SK Hynix invested a cumulative total of about 860 billion Korean Won in HBM-related R&D and about 1.5 trillion Korean Won in facilities and equipment. A significant portion of these investments occurred during years when the market was sluggish and HBM's commercial prospects were completely unclear.

What happened during this period?

HBM2 failed to meet performance expectations and had to be redone, resulting in a revised version called "HBM2 Gen2." The HBM team internally became known as "the department nobody wanted to join." Key personnel were rotated out, and morale was low. AMD's flagship graphics card, the R9 Fury X, used first-generation HBM in 2015, but the market reception was lukewarm because it was too expensive; no one bought it.

What made Hynix even more panicked was the period around 2016-2017. Broadcom approached Samsung, hoping Samsung would supply HBM2 for Google's second-generation TPU. If Samsung could meet 100% of the demand, Broadcom promised them exclusive supply rights. This should have been the crucial moment for HBM to first enter the data center.

Unexpectedly, Samsung dropped the ball.

A report from the Korean *JoongAng Ilbo* recounted the chaos at the time: The Google TPU project involved three companies: Broadcom (design), Samsung (memory), and TSMC (foundry). Samsung's HBM had memory issues. Engineers reported upwards that TSMC refused to let them enter the factory for inspection. The three companies blamed each other, the problem remained unresolved, and the delay lasted up to six months. "This kind of deadlock was common between 2016 and 2017," a knowledgeable executive later recalled.

Google later gradually leaned towards collaborating with SK Hynix. The biggest beneficiary of HBM's first real deployment in the data center was not Samsung.

But as all this was happening, the world had no idea that HBM would become the most critical bottleneck of the AI era.

In an interview for *Super Momentum*, Chey Tae-won later simply said, "We were standing at a crossroads."

He didn't elaborate on why he persisted. But in hindsight, the logic might be this: Hynix's fate had already brushed against bankruptcy in 2001. It didn't have diverse businesses like Samsung to spread risk; memory was the only thing it could do. Either it became number one in the world at this one thing, or it would live forever in Samsung's shadow.

So, "betting on HBM" wasn't a choice for Hynix; it had no choice.

2022: Jensen Huang Hands Over a Match

In June 2022, SK Hynix began mass-producing HBM3. That year, the first batches of HBM3 chips were installed on a GPU called the H100, made by a company then worth around $300 billion, considered "important but not world-changing" in both gaming graphics and data centers: NVIDIA.

In November, OpenAI released ChatGPT.

What happened next is well known. The demand curve for AI computing power transformed from a gentle slope into an almost vertical rocket trajectory. Every GPU used for training large models needed HBM as its closest "data porter."

At that moment, the hand Hynix had been playing for 14 years was finally turned face up.

By the second quarter of 2025, Hynix had captured 62% of the global HBM market share. Samsung had fallen to 17%, even trailing behind the latecomer Micron (21%).

By the full year 2025, Hynix reported an annual operating profit of 47.2 trillion Korean Won, while Samsung Electronics reported 43.6 trillion Korean Won. This was the first time in Hynix's history that it surpassed Samsung in annual profit.

During the Taipei Computex event in August 2025, NVIDIA CEO Jensen Huang visited the SK Hynix booth and left a handwritten note on the display board: "JHH LOVES SK HYNIX!" This photo was widely circulated by Korean media. In engineering culture, there is no more direct form of official endorsement.

And SK Hynix's engineers jokingly gave a new full name to HBM, saying it actually stands for "Hynix Best Memory".

Where Did Samsung Actually Lose?

So, the question arises: Samsung, the company that once crushed Japanese DRAM factories and pushed Micron into a corner, where did it actually lose?

Samsung missed early positioning in HBM, made the wrong technology choice (NCF packaging vs. Hynix's MR-MUF packaging), and its HBM3E kept failing NVIDIA's certification tests...

These are all facts, but they aren't the root cause. The root cause is something more awkward and ironic: Samsung was too successful, so it couldn't afford to lose and couldn't afford to gamble.

Compare the two companies at that pivotal point in 2008. Hynix had just crawled out of bankruptcy trusteeship. The entire company had only one business line: memory. It had no diversified cash cow to fall back on.

It bet on HBM not because its vision was exceptionally clear, but because it had no other option. It had to seize any chance that could help it shed the label of "Samsung's little brother."

And what about Samsung?

In 2008, Samsung was on the eve of its peak. The mobile phone business was about to take off with the Galaxy series. The semiconductor business was number one globally in both DRAM and NAND. The display business was about to land massive OLED orders from Apple's iPhone. Its cash flow was incredibly strong, its business portfolio was enormous, and the stakeholders it needed to balance were extremely complex.

For a company like this, what was HBM in 2008? It was a high-risk bet with a tiny market, a long payback period, and one that conflicted with another technology path (HMC) it was championing. Any rational finance committee would have refused to approve going all-in on this.

This is the classic "Innovator's Dilemma": Successful large companies are forever constrained by their own success. The markets they have already won are too big, too important, and need too much protection. They cannot, like a cornered challenger, place their entire bet on a seemingly unreliable new direction.

The deeper irony is that Samsung wasn't blind to HBM. It started investing in HBM-related research as early as 2011 and even mass-produced HBM2 first in 2016. But each time, Samsung didn't go all-in. Its energy was simultaneously spread across a dozen battlefronts like HMC, GDDR, LPDDR, and enterprise SSDs. When Hynix's HBM team was "marginalized but still grinding away," Samsung's HBM team was also "marginalized," only no one was there to grind for them.

By 2024-2025, Samsung finally realized it had to go all-in on HBM, but it was too late. A gap in technical processes had already formed, and the moat in customer relationships had been built jointly by NVIDIA and Hynix.

Jun Young-hyun, Vice Chairman of Samsung Electronics' semiconductor business, said in his 2026 New Year's address: "Our customers are telling us that Samsung is back."

"Is back"—these three words are themselves an admission.

Two Questions

What does the story of SK Hynix signify? At least two questions are worth pondering.

First, why do similar stories tend to happen more easily in South Korea than elsewhere?

SK Hynix's success didn't come from nowhere. It is supported by a unique industrial soil. Although the Korean chaebol system has been criticized for decades, it objectively allows a company, driven by the will of a single decision-maker, to bet on a gamble with a 20-year return cycle and to keep providing capital for the middle 10 years when no commercial prospects are visible.

When Chey Tae-won acquired Hynix in 2012,