The 120-Nanometer Leap
During the height of the smart war, when Intel and AMD had just released the 586, Yoo Jae-won unveiled copper interconnect technology alongside Dr. Lisa Su, dramatically boosting chip operating speeds. The performance gap with existing chips was so vast that a separate certification called HPC was created. Years later, semiconductors had become so ubiquitous that the certification lost all meaning, yet there had been a time when the mere presence of the HPC mark commanded a premium. This time, Yoo Jae-won intended to work with Dr. Lisa Su once more, though the approach differed from before. The technology itself was different, of course, but the greatest distinction lay in whether the semiconductor technology would be disclosed at all.
During the HPC era, ID Group had possessed no semiconductor production lines of its own. That was why the copper interconnect technology had been made available to Intel, Mirae Electronics, and even Ilsung Electronics—at remarkably low prices. In truth, the technology had emerged largely because computers had been so slow that Yoo Jae-won himself grew impatient. This time was different. With ID Electronics’ own semiconductor division and production lines, he had no intention of releasing the technology as freely as before. Moreover, the new technology could not be applied simply by disclosure. It required full optimization across the entire semiconductor manufacturing process, and the chemicals involved were so different that, in some cases, entire production lines would need to be rebuilt.
Yoo Jae-won had already sketched the larger picture: maximize profits first through ID Electronics, then license the technology individually to partners such as Intel, AMD, ATI, and NVIDIA. Before long, Yoo Jae-won, Tiffany, and Dr. Lisa Su reached the pilot production line. As the world’s largest memory semiconductor fabrication plant, even the pilot line was enormous—not a laboratory-scale facility, but an actual production line lifted straight from the main plant. The equipment alone was worth hundreds of billions of won. ID Electronics semiconductor team researchers were already waiting, their excitement barely contained. Dr. Lisa Su’s stature in the industry was immense. A small misunderstanding existed: the public had come to regard Dr. Lisa Su as the principal developer of copper interconnect technology, when in reality Yoo Jae-won had led the effort and she had played a supporting role. Semiconductor experts assumed that even if Yoo Jae-won was a genius, his talents were limited to programming and did not extend to semiconductor materials. Yoo Jae-won had deliberately allowed that perception to stand—until the moment he began briefing the semiconductor team on the technology he had prepared.
“The technology to be applied to the MAP 3 chip consists of three-dimensional transistor fabrication and ultra-precision etching using high-purity hydrofluoric acid gas.”
Yoo Jae-won moved straight to the point, projecting the specifications of the MAP 3 that would be mounted in the next-generation smartphone: an 800 MHz dual-core mobile processor built on 120-nanometer technology, integrated with ATI Radeon architecture for graphics processing. The goal was to deliver at least three times, and up to four times, the performance of the MAP 2 currently inside the T-Touch Phone.
“Excuse me, is this actually possible?” a senior researcher asked cautiously. ID Electronics was currently running at 240 nanometers for memory semiconductors. The idea of halving that figure in one leap naturally struck everyone as impossible. Microfabrication development was already underway internally—raising density from 240 to 180 nanometers—but two years of work had only produced modest results. Now they were being asked for 120 nanometers, and not for relatively simple memory chips, but for a mobile processor whose complexity far exceeded that of memory devices. Success would place them ahead of Intel and AMD in process technology. Applied to memory semiconductors, the impact would be enormous: halving the feature size from 240 to 120 nanometers would quadruple the density per unit area and potentially quadruple the number of chips produced per wafer.
“Heh, you’ll see soon enough,” Yoo Jae-won replied with confidence. Though the master plan had lost much of its original meaning, the future technologies collected while preparing it were genuine. The three-dimensional transistor technology and high-purity hydrofluoric acid etching would remain highly efficient tools until silicon-based semiconductors reached the end of their lifespan. Adding EUV lithography to the mix would allow single-digit nanometer processes without difficulty.
“Ah, does anyone here have contact with Professor Lee Jong-hyo of Seoul National University?”
The researchers exchanged glances. Everyone present was among Korea’s top semiconductor experts, all graduates of Seoul National University’s electrical engineering department, where Professor Lee Jong-hyo’s classes had been mandatory. “He’s said to be an authority on three-dimensional transistor technology. I’d like to propose an industry-academia collaboration.”
In truth, Yoo Jae-won already possessed the complete technology and did not strictly need Professor Lee’s assistance. Yet Lee Jong-hyo had been the originator of three-dimensional transistor concepts. The gap in sophistication between the professor’s work and the knowledge in Yoo Jae-won’s mind was vast, but bringing the original creator aboard would ease the burden on his conscience.
“Yes! I know him!” the senior researcher answered at once. A Seoul National University electrical engineering alumnus who had studied under Professor Lee, he still maintained close contact. He had long felt sorry that his former advisor received little institutional support and struggled to secure government funding. If this unexpected opportunity could be seized, his mentor might become a second Dr. Lisa Su.
“Really? When was the last time you spoke?”
“A week ago!”
“Wow, you two must be quite close. Then I’ll leave it in your hands.”
“Yes! I’ll contact him right away.”
“Then Senior Researcher Na, please handle Professor Lee Jong-hyo. Shall we begin the actual work?”
After assigning the task, Yoo Jae-won turned to the pilot production of MAP 3 in earnest. “Dr. Lisa Su, if you would.”
The briefing on MAP 3 architecture began. A new legend surpassing copper interconnect technology was about to unfold in Daejeon, Republic of Korea.
“Jae-won! Merry Christmas!”
“Merry Christmas to you too!”
On Christmas Eve, Yoo Jae-won and Tiffany exchanged special greetings and the gifts they had prepared. He gave her a bouquet and a necklace; she gave him a warm sweater. The atmosphere was perfect. The night view of Seoul from the penthouse atop ID Global Headquarters was spectacular, and having the person he loved beside him made the intimate Christmas exactly what he had always imagined. A soft meow sounded—there was also a cat. Still, time passed far too quickly.
The oversized LCD television that dominated the living room was broadcasting the latest from America. The U.S. presidential election had unfolded almost exactly as Yoo Jae-won had predicted. The early blow of the IT bubble collapse had actually benefited the Democrats. Al Gore, heeding Yoo Jae-won’s advice, had aggressively used SNS and the internet in his campaign, securing solid support from younger voters and ultimately carrying Florida despite the controversies. The recount had produced many invalid ballots, yet Gore’s margin, though narrow, held. Because his victory was already assured regardless of Florida, the dispute never escalated. The Bush camp requested another recount merely to avoid the label of total defeat, but the second count only widened Gore’s lead. The Democratic Party was in full holiday-party mode; it had been a long time since they had managed to extend their hold on power. On top of that, President Clinton had received the Nobel Peace Prize for his role in establishing peace on the Korean Peninsula. It was truly the golden age of the Democratic Party.
The television delivering the news was itself extraordinary. Its elegant design harmonized with the penthouse interior, and its 42-inch screen offered perfect 1080p full-HD resolution—an advanced model produced by ID Display, the largest LCD television in the world. It was the kind of product money alone could not buy; only the finest unit from ID Display’s laboratory currently graced Yoo Jae-won’s living room. Mass-production preparations were underway, though the target remained late 2004. The technology was ready, but the manufacturing infrastructure was not. Even then, the sets would be expensive, yet Yoo Jae-won believed that by the mid-to-late 2000s every household would have one. Meanwhile, 30-inch 720p models were already in steady production and prices continued to fall; by summer or autumn of 2001 they were expected to dip into the low one-million-won range. Broadcasters planned to begin full-scale HD transmission around the 2002 World Cup, so everything was proceeding smoothly. Xbox sales benefited as well. As Yoo Jae-won had foreseen, the more LCD monitors proliferated, the more clearly the performance gap between Xbox and PlayStation 2 became apparent, steadily increasing the number of gamers choosing Xbox.
Not everything went according to plan, however. The vision he had announced at the beginning of the year had encountered the greatest delay. Whenever ID Group unveiled a vision, concrete change followed. SNS had been incomprehensible at first, yet now everyone used it fluently—sometimes too fluently, posting thoughts better kept private and suffering the consequences. Celebrities trying to manage fans on SNS frequently caused scandals with ill-considered posts. This time was different. In January Yoo Jae-won had spoken of system integration and AI, yet with only days left in 2000, visible progress remained elusive. The core AI algorithms were already complete and were undergoing silent machine learning under Yeong-sik’s supervision. Unlike SNS, which had quickly taken concrete form, the only immediately commercializable AI application was the filtering system. The real bottleneck was system integration. Yoo Jae-won intended to unify all of ID Group’s computer systems and then layer AI on top, placing every employee’s computing activity within the machine-learning domain. The ultimate aim was for AI to monitor the group’s cash flow and eventually replace repetitive tasks. After nearly a year, integration of the U.S. subsidiaries was largely complete, but work in Korea was still ongoing. Unlike the United States, Korea had seen widespread acquisitions and mergers, and the merged companies each used different systems. The difficulty was extreme, though not beyond resolution given enough time.
Another matter was also taking longer than expected.
“Ah, how are the wafer yields this time?”
Tiffany’s question struck exactly at the issue. The purpose of this trip to Korea was to begin in-house production of MAP 3, the centerpiece of the smartphone war. Dr. Lisa Su’s team had completed the MAP 3 architecture, and Yoo Jae-won had introduced two new technologies to enhance semiconductor performance. They had begun pilot production with confidence, yet results had fallen short of expectations. It was as though they possessed a perfect recipe and followed it precisely, only to find the dish did not taste as intended.
“Another failure, apparently.”
The latest wafer had yielded disastrously—not because the chips failed to reach the target speed, but because so few chips functioned at all.
“At least Professor Lee Jong-hyo seems to have found a clue.”
“Really? That’s a relief.”
Yoo Jae-won had invited Professor Lee because he felt uneasy relying solely on his own knowledge. The synergy between the professor and Dr. Lisa Su proved far greater than anticipated. While Yoo Jae-won supplied the textbook solution, Dr. Lisa Su and Professor Lee were solving the countless practical problems that arose on the line. As Tiffany said, it was fortunate indeed. Still, their method amounted to endless cycles of trial, error, and refinement. Preparing for the smart war through nothing but repetitive drudgery felt strangely anticlimactic.
While chatting quietly with Tiffany, Yoo Jae-won suddenly heard his own name on the television.
“Hmm? You got invited to the Blue House?”
Tiffany’s question made him shake his head. He had never heard anything about it.