The Second Opportunity
They say that every person gets at least three major opportunities in life. Yoo Jae-won realized the same principle applied not only to individuals but also to companies and organizations while observing AMD. Even the perennial runner-up in the CPU world had been given three chances to surpass the undisputed leader, Intel.
If one were to label AMD’s three opportunities by the codenames of its products, the first would be the Athlon era, the second the Zen series, and the third the Cold Fusion period. Among them, Lisa Su had been the architect of the second opportunity—the Zen series. Especially during the time when Zen versions 1 through 6 were released, AMD could confidently claim it was enjoying its greatest golden age in history. The company had overtaken Intel in CPU market share and maintained that lead for several years.
“Lisa Su, huh.”
What made her even more remarkable was that she was both CEO and a hands-on developer. That was why people naturally called her “Doctor” rather than “Madam President.” She herself identified more strongly as a developer than as an executive, and she held a Ph.D. in electrical engineering from MIT.
“Hmm. So if I get involved, her career trajectory might change as well.”
Yoo Jae-won had been about to send her an email when he suddenly hesitated. The message would express ID Group’s interest in her master’s thesis and propose joint follow-up research. If she showed interest, he was prepared to send a plane ticket or even fly out personally to recruit her.
“This is getting complicated.”
Intervening in the career of such an important figure could trigger unpredictable aftereffects. What if, armed with experience gained at ID Group, she later became CEO of Intel? That would be nothing short of disastrous for AMD. Because Yoo Jae-won had always felt more favorably toward AMD than Intel, his hesitation grew.
After all, in his previous life, when money had been tight, AMD had consistently delivered good value. Without AMD, he would have been forced to buy mediocre CPUs at premium prices. Thanks to AMD’s competition, he had been able to acquire multi-core processors at reasonable cost. The same held true in the quantum computing era. AMD’s Cold Fusion series had been the only real competitor to Intel’s quantum CPUs. While absolute performance still lagged behind Intel, the far lower price point had made decent performance accessible to countless developers.
“Let me think for a moment.”
The moment of contemplation stretched past ten minutes. Inside Yoo Jae-won’s mind, countless scenarios played out simultaneously. Without a quantum computer, the simulations could never be perfectly accurate, yet decades of accumulated experience allowed him to estimate outcomes ranging from best to worst case and assess whether they could be controlled. Roughly thirty minutes later, he reached a conclusion.
“Let’s do it.”
The answer was simple: one should not refuse to make soybean paste just because one fears the maggots. Whether Lisa Su would accept his proposal remained unknown. If she did, excellent. If not, proper compensation for citing her thesis would suffice. AMD itself could be treated as a separate matter. The Zen series would not appear for another twenty-four years. Even assuming two or three years of research and development, any repercussions would not surface for roughly twenty years—an impossible variable to predict even with GS-X, the twenty-first century’s most advanced quantum computer.
More importantly, with Yoo Jae-won’s full-scale intervention rapidly pushing global history into unknown territory, the likelihood that something would happen to Lisa Su before she could become AMD’s second savior was far higher. She might even achieve greater success than before. The Win-Tel alliance that had once dominated his previous life was already gone, and AMD CPUs were already installed in Nextcomcast’s massive data centers. Consequently, optimization for the enterprise version of Android OS was also likely to be based on AMD architecture.
“If things go wrong, I can always buy them out.”
The master plan naturally included expansion into the semiconductor sector. Therefore, if AMD’s situation deteriorated contrary to expectations, acquisition remained entirely feasible.
“I hope she gives a positive reply.”
Yoo Jae-won composed a polite invitation to Lisa Su and pressed send with hope in his heart.
Silicon Valley was a small town. Within just a few days, rumors that ID Group was entering the semiconductor business spread like wildfire. While the plan to build an LCD factory had been officially announced, the news of semiconductor technology development and design leaked rapidly even though it had not been formally disclosed.
The word that ID Group was actively recruiting semiconductor experts struck Silicon Valley like a thunderbolt, raising expectations among many. The most recent and most definitive success story in Silicon Valley was ID Group. Once dismissed by many as a fleeting bubble, the company was now beyond doubt. For such a group to venture into semiconductors implied that it had already achieved positive internal results. Everyone knew ID Group possessed enormous capital, so massive investment was expected, stirring excitement among those hoping to ride the wave.
Intel and AMD reacted in unusual ways. Naturally, their responses differed subtly. As the established leader of the CPU industry, Intel adopted an attitude of welcoming any new challenger, while simultaneously offering “advice” that semiconductor design and manufacturing were no easy fields and that thorough mental preparation would be required. The implication was that the ever-successful ID Group would this time receive a valuable lesson in failure. In any case, Intel remained confident that ID Group would never surpass its own semiconductor technology.
AMD, on the other hand, reacted with genuine alarm. Not long ago, ID Group had purchased twelve thousand of AMD’s highest-performance CPUs at once to build a massive data center. AMD had been thrilled. With the rapid spread of Pentium-class computers and the internet, demand for server computers to provide internet services was bound to surge. The fact that ID Group—the very creator and fastest leader of internet technology—had chosen AMD systems for its data center suggested that countless other companies would follow suit. Now, if ID Group turned to in-house development instead, AMD would be left chasing after nothing.
The anxiety was not mere worry. Following the news that AMD products had been adopted for ID Group’s data center, AMD’s stock price—which had been steadily rising—reversed and began to fall. AMD’s management, feeling the heat, took action. They repeatedly called ID Group’s secretariat until they secured a meeting with Yoo Jae-won.
“Ah, a phone call would have been enough. I’m sorry you had to come in person.”
Yoo Jae-won’s apology stemmed from the date. The meeting with AMD’s president, Jerry Sanders, had been scheduled for Christmas Eve. While not quite as significant as Christmas Day itself, Christmas Eve was still a meaningful occasion—one Americans typically spent with family or loved ones. When setting the date, ID Group had said any day was fine, and AMD had wanted to meet as soon as possible, resulting in December 24th.
“Not at all. If anything, I should apologize. I was so concerned about the future of AMD’s employees and customers that I ended up disturbing your Christmas, Mr. Chairman.”
Jerry Sanders looked like a refined gentleman. His hair was completely white, and so was his beard. The black suit suited him perfectly. Yet contrary to appearances, Jerry Sanders was known by the nickname “Tough Guy” and was extremely aggressive—not in the sense of throwing punches, but in his preference for aggressive brand positioning and marketing. The fact that AMD, perpetually behind Intel in every metric, had still managed to hold on to second place was entirely thanks to Jerry Sanders’ innate instincts. It was also his strategy that had led AMD to adopt Performance Rating (PR) numbers instead of actual clock speeds, and the aggressive pricing policy had begun under his leadership.
The K5 PR66 model that Yoo Jae-won had ordered twelve thousand units of actually ran at 50 MHz. AMD had promoted it by claiming that because its architecture differed from Intel’s, the real-world performance matched Intel’s Pentium 66 MHz model, hence the PR66 rating. That claim was both true and false. The performance parity only appeared in specific applications like benchmark programs, not across all workloads. Still, consumers chose AMD because it was far cheaper than Intel. AMD had even used Yoo Jae-won’s decision to build a data center with its CPUs as a marketing point for enterprise customers.
That was why Jerry Sanders’ attitude toward the news of ID Group entering the semiconductor field was completely different. It was no surprise that the AMD president had sacrificed Christmas Eve to meet Yoo Jae-won.
“So it was a misunderstanding!”
Jerry Sanders’ voice rose. Yoo Jae-won proved far more reasonable than the AMD president had imagined or prepared for. Though outwardly a young man, he possessed the mindset of someone over sixty and was favorably disposed toward AMD. Most importantly, the semiconductor business Yoo Jae-won envisioned was not the kind AMD had feared.
Jerry Sanders and Intel had assumed ID Group would produce its own CPUs and enter the retail market. What Yoo Jae-won actually intended, however, was to sell technology. The copper interconnect process he planned to develop was a manufacturing technology applicable to all semiconductors. ID Group had no intention of operating semiconductor fabrication plants, which were cumbersome, capital-intensive, and environmentally damaging. Therefore, the only remaining option was to sell the technology.
Meanwhile, low-power CPUs for mobile devices were products that neither Intel nor AMD had yet developed. ID Group would design them in-house and outsource actual production to OEM manufacturers such as TSMC or UMC. All units produced this way would be consumed internally by ID Group, though excess units could be sold to other companies if demand existed. Retail sales to individual consumers, however, would never happen.
“A copper interconnect process…”
As AMD’s president, Jerry Sanders immediately grasped the value of the technology Yoo Jae-won described.
“Don’t tell me you’ve already completed the formula?”
Jerry Sanders, who had just begun to relax, immediately showed interest. Naturally, Yoo Jae-won shook his head. This was not the right moment to reveal what he already knew.
If Lisa Su had responded positively to his proposal, he could have given an affirmative answer here. But there had been no reaction whatsoever. The email remained marked “unread,” and calls to her MIT lab went unanswered. It seemed her entire research team had already left for Christmas vacation. He had considered looking up her personal contact information, but that proved difficult. In the future she would become an enormous figure, but right now she was merely one of countless researchers who had just earned their doctorates.
He hoped she would reply before Bill Clinton’s presidential inauguration at the latest.
“I’ve only had the idea so far.”
Consequently, Yoo Jae-won’s answer was less than satisfying, and Jerry Sanders’ previously expectant expression quickly settled.
“Ah, still, it’s a remarkable idea.”
He did not forget to offer lip service. Although the project was only in its earliest stages, ID Group’s financial resources had long since surpassed AMD’s. Moreover, countless people had already learned the hard way that it was dangerous to judge the enigmatic Yoo Jae-won by conventional standards. Jerry Sanders remembered how Gates and Steven, who had once dominated the PC operating system market, had grown complacent and vanished overnight. There were rumors they were tinkering with new projects in an attempt to stage comebacks, but those rumors were from a year ago and had not been updated since, so they were effectively finished.
“Then may I ask how you intend to implement the technology?”
With his keen intuition, Yoo Jae-won immediately recognized that Jerry Sanders was probing him.
“Heh, that’s for later. Once the technology is complete, we’ll hold a major showcase, so you can see it then. I won’t keep you waiting long either. Hmm, probably by next summer at the latest?”
Jerry Sanders’ expression grew subtle. The dissonance arose because he was forcibly maintaining a composed face to hide the “doubt” that had momentarily surfaced. He acknowledged that Yoo Jae-won was an extraordinary individual, but that assessment was limited to the software field. Semiconductors—the foundation of computer hardware—were an entirely different domain. Next summer? That was at most seven months away. Seven months was an extremely short period in semiconductor development. Creating a groundbreaking manufacturing process in that timeframe was absurd. If the person sitting before him had been anyone other than Yoo Jae-won, Jerry Sanders would have accused him of fraud and walked out.
“I can see all your concerns. When the time comes, we’ll hold a proper showcase, so don’t worry. Besides, I’m not asking you for an entrance fee right now, Mr. Sanders.”
Yoo Jae-won was the foremost expert in packaging technology. Even ID Office, which had once lacked presence, had instantly captured the market by hastily organizing the global Security Challenge event. ID Group’s approach to physical packaging was equally distinctive. After the phosgene incident, both ID Group and TG had gone to great lengths in environmental protection. When creating packages, they minimized the use of plastic and used recycled cardboard. They had long since perfected the art of folding that cardboard to provide not only shock absorption but also a sense of luxury.
The packaging of the recently released Shell Book was a prime example. It paid homage to IT device packaging from the mid-twenty-first century, and its high level of completion naturally meant significant production costs. Yet Yoo Jae-won had pushed forward anyway. Sometimes irrational decisions produced excellent results. The high-end eco-friendly packaging had instilled in consumers the perception that products bearing the ID mark were genuinely different, turning them into objects of desire and generating substantial demand.
Furthermore, all packaging was manufactured at the factory in Yeoju, South Korea. The facility had become so busy that it produced hundreds of thousands of units per month, employing over two thousand people and contributing significantly to regional development. Assembling the cut cardboard packages was difficult to automate, so human labor remained essential. As a result, people working at the packaging factory were a common sight in Deokjin-ri and Yeoju, which in turn boosted Yoo Jae-won’s popularity.
In any case, Yoo Jae-won held strong convictions about packaging. He also understood the value of the technology he was about to unveil better than anyone. Therefore, preparing a grand showcase was only natural.
After Jerry Sanders left, Yoo Jae-won found himself with time on his hands. The story of ID Group’s new semiconductor venture, which had briefly shaken Silicon Valley, faded into the background as the Christmas and New Year season arrived. That was how news worked. Unless one was directly involved, stories came and went in an instant. The working-level staff, however, remained frantically busy despite the year-end holidays. True to its massive yet agile nature, ID Group had moved immediately after Yoo Jae-won mentioned LCD and semiconductors. Hwang Jae-hong of the Korean branch had gone down to Daejeon to scout industrial park sites for the factory, while in Silicon Valley they had openly begun recruiting semiconductor experts. Buildings to serve as research labs had already been secured. They were even receiving quotes from TSMC and UMC for contract manufacturing.
Yoo Jae-won’s free time stemmed from the fact that while all preparations were proceeding smoothly, the one person he needed to align with had yet to respond.
“By the way, Lisa Su, you read the email but why haven’t you replied? Are you really going to stay silent until next year?”
The Lisa Su Yoo Jae-won knew was the very embodiment of boldness. Jerry Sanders and AMD’s successive presidents and CEOs all seemed to require that larger-than-life personality. That was why she had earned the nickname “General” in addition to “Doctor.” He could not understand why she was being so passive now.
Yoo Jae-won sent one more email and then deliberately stopped thinking about it. For now, he simply wanted to enjoy the modest Christmas holiday. More precisely, his plan was to sit in front of the computer and play games throughout Christmas and the following weekend—a plan worthy of a true homebody.
Game selection was important, and Yoo Jae-won already had his answer: Doom 2. Released in time for the Christmas season, Doom 2 did not disappoint. Despite the warning label about its extremely violent content, both children and adults played it enthusiastically. Yoo Jae-won had cleared the game on the day it came out and had started multiplayer the day before. There were multiple maps designed exclusively for multiplayer, and the host could freely set the rules. At Yoo Jae-won’s suggestion, a modern matchmaking system had been implemented that assigned ranks based on match records and matched players accordingly, allowing for long-term enjoyment.
The issue was that comfort levels varied greatly depending on hardware performance and network environment—an unfairness external to the game itself. Naturally, Yoo Jae-won, equipped with top-tier specs that could not be improved further and the latest ADSL connection, held an overwhelming advantage. Moreover, he was thoroughly familiar with FPS games like Doom 2 and rules such as Team Deathmatch and Free-for-All, so his rank climbed rapidly.
Thanks to Doom 2’s enhanced ranking system, far more people enjoyed multiplayer than single-player. Although internet penetration was still low, anyone who used the internet and bought Doom 2 played multiplayer without exception. The computer industry, already enjoying peak season during Christmas, was practically screaming with joy. Doom 2’s success had increased applications for internet service, and people were upgrading their computers and buying better 3D accelerator cards every day in order to win at multiplayer.
December 27th. It was the final day of the Christmas holiday, yet Yoo Jae-won’s appearance was no different from the 25th. Eating, sleeping, and gaming—as if that were his sole mission on earth.
—Triple Kill!
A cheerful sound burst from the high-end speakers. It was the alert that played when a player killed three enemies in quick succession.
“Heh, triple kill is nothing.”
While most gamers would have jumped for joy at such an outstanding play, Yoo Jae-won had already achieved it seven times today—eight including this one—so his reaction remained indifferent.
“Ugh! I died.”
Instead, he lamented failing to finish off the remaining enemies before dying himself. He had seen plenty of triple kills and pentakills, but he had yet to achieve an “all kill” by eliminating the entire opposing team. Since the match had ended, he would have to wait for the next room to open.
As he stretched his stiff body, a loud knocking sound echoed. Thinking a new game had started, he quickly turned his eyes to the monitor, but that was not it.
“Huh?”
It had not been his imagination. Moments later, the sound of a ringing phone followed. It was the intercom connected to the security post outside.
“Yes, what is it?”
The device was designed to notify him immediately of any issues outside the house, so Yoo Jae-won picked up the receiver right away. After hearing a few words from the other end, he sprang to his feet and rushed to the entrance. Then he flung the door open.
Standing there, alongside the security guard, was Lisa Su with an awkward expression on her face.