Intel's Handshake and the Hovering Prototype
Among the companies that had entered the business hall, only Intel had yet to sign a contract. “Just give me a little more time and I’ll come running right away!” The team leader, who had watched everyone else head off for jajangmyeon like a child left sucking his thumb alone, refused to give up. “Of course. The doors of ID Group are always open.” Yoo Jae-won nodded at the development team leader’s words and replied. Even though the processing power of memory, 3D accelerator cards, sound cards, and the like had improved dramatically thanks to the Coppermine process, the crowning touch was still the CPU. Completing HPC Class meant unifying everything, right down to the CPU, under the Coppermine process. In that situation, Intel’s participation—the largest player in the CPU industry—was an essential element of Yoo Jae-won’s grand vision.
Having wrapped up business perfectly, Yoo Jae-won’s expression grew far more relaxed. The numbers worked out too. Though the event lasted only a single day, preparation costs alone had run into the millions of dollars. Building the stage and decorating the interior had been work enough, but rushing partner companies to mass-produce HPC Class components for the main-stage demonstrations and experience zones had been no small task. Naturally, that pressure had not come for free; it had been backed by ID Group’s emergency orders. In other words, every HPC Class component shown at the event had been hand-built at enormous expense.
Although admission had been charged in euros, the price was so low that the only real cost had been renting the convention center. Thanks to strong sales of the Coppermine license from day one, the event had already turned a profit. Moreover, the license’s billing structure meant that the more semiconductors sold, the larger the royalties ID Technology would receive. A somewhat complex calculation was required, but the figures were roughly ten dollars per CPU, five to seven dollars per 3D accelerator chip, two to three dollars per DSP chip for sound cards, and under one dollar for I/O expansion chips such as Ethernet. With a twenty-year patent term, ID Technology had effectively become a company that could never go bankrupt even if every new venture it launched failed.
The only potential risk was semiconductor manufacturers underreporting production volumes. License fees were determined by the number of wafers used, and it was impossible for ID Technology to station staff at every partner factory to count them. Still, Yoo Jae-won was not particularly worried. The reason was simple: he controlled the computer operating system himself. Those components did not function independently. Only after they were assembled into a complete system on a motherboard and an operating system was installed would a computer actually run. The adware version could collect users’ system configuration data, which could then be aggregated to estimate each chip’s production volume. If a manufacturer’s self-reported figures diverged significantly from the collected data, an audit could begin immediately.
While some users might refuse to share system configuration information, that was precisely why HPC Class had been created. It was not enough for the HPC Class mark to appear on the component box. After booting, the mark could also be verified in the system manager’s component list—provided the HPC certification program had been run. That certification program itself functioned as a statistical tool. Computer users were sensitive even to a single small mark. Anyone who had purchased an HPC Class component would naturally check whether the mark appeared correctly in the operating system. If a company failed to pay the proper license fees, the statistical data gathered this way could be used to confront them, and if they still refused to comply, the license could simply be revoked.
Having successfully concluded business, Yoo Jae-won returned to the main stage in a lighthearted mood. The area had already become a festival ground. Contrary to the expectations of some executives who had worried that watching others play games would be boring, cheers erupted whenever a ranked player scored a kill, and sighs of disappointment followed when a supported player fell. The ranked players currently competing were truly delivering the performances of a lifetime.
“Whoa! Railgun! Railgun respawn!”
“That player must have a timer in his head! He grabs the railgun the moment it spawns and fires—double kill!”
A thrilling announcement rang out over the main-stage speakers.
“The other ranked players will know by now! Once that player picks up the railgun, there’s no stopping him!”
“There it is again! One-shot! Two kills!”
The caster and commentator brought in by Nextcomcast, claiming they could broadcast any match, heightened the tension of the game even further. The star of the show was ID Fatalist, the youngest ranked player. Armed with a weapon that dealt massive damage but required a long recharge after each shot and was difficult to aim, he swept through the competition. The fact that a single railgun shot had raised his kill count by two was no bug. The railgun’s projectile was armor-piercing, and two players had unfortunately lined up in its line of fire. It was the kind of super play that was hard to pull off even after dozens of matches at home, and he had already achieved it twice.
Fatalist was not merely skilled physically; his strategic judgment was also excellent. In the solo tournament, played in a 15-to-1 format where everyone else was an enemy, players had to finish in the top eight to advance. Normally this encouraged passive play that could make the match dull, but Fatalist had turned that logic on its head by taking the initiative. To prevent boring games, kill points had been set generously, directly tied to bonuses, and he had boldly pursued aggressive play to claim them.
“That’s it! Take him down!”
While Fatalist’s flashy play was impressive, even his supporters had grown heated. No one had expected Tiffany—dressed in a gown and perfectly made up—to swing her fist through the air while cheering “One more kill!”
“Huh? When did you get here?”
Startled upon noticing Yoo Jae-won approaching, she jumped. “A little while ago. Anyway, is the match fun?”
“Huh? Of course! Didn’t you just see that super play? It was incredible. Sit down already—you’re blocking the view!”
Yet Tiffany did not try to hide her excitement now that she had been caught. She had never been the demure type to begin with, and with her attention fully absorbed by the match, Yoo Jae-won was momentarily pushed aside. That was very much like her. At least he was better off than Gilbert, who was completely invisible to her. The man’s soul had already left his body and was glued to the main screen. Yoo Jae-won gave a wry smile and quietly took the empty seat to Tiffany’s left. He had hoped to chat until the awards ceremony after the match, but it seemed he would have to wait until the game ended. Still, one thing was certain: today’s event, named the HPC Expo by Coppermine, had been an unqualified success.
A few days later.
—Jae-won! Inquiries about the New Egg 2 are pouring in. It’s chaos over in the States.
Lee Yong-kwon’s ID Talk message arrived while Yoo Jae-won was working in his study as usual.
“Just as we expected.”
User loyalty to the Egg series was extraordinary compared with other computer manufacturers. Large PC makers like Compaq and Dell had been selling massive volumes for years, yet none had achieved the same level of devotion as the Egg line. That loyalty stemmed from the Egg series’ unique identity. It was the ultimate PC and also the ultimate interior piece. Its sophisticated design looked at home anywhere, and its performance was flawless. Its only drawback was a somewhat higher price than ordinary computers, but that very fact gave it the aura of a luxury item that not just anyone could own.
—Still, this one’s different. The aluminum case and frame really hit the mark.
The aluminum used in the New Egg 2 was unmistakably futuristic. Rather than a cheap, shiny finish, the matte silver coating had been the key, according to reviews. Moreover, the sight of the seemingly thin and light aluminum frame effortlessly suspending a large Trinitron monitor in midair was visually striking. Of course, the Egg PC still delivered top-tier performance, and the addition of HPC Class certification this time created tremendous synergy. Naturally, the HPC Class brand itself had gained massive recognition from the event. ID Technology’s name appeared in every semiconductor or IT-related news story worldwide. Once again hailed as a miracle, the company was praised for instantly multiplying the performance of existing semiconductors several times over. While ID Group’s capable PR and marketing teams deserved credit, the sheer destructive power of the Coppermine process had made it possible.
“As you know, an immediate launch is impossible. Let’s just take reservations for now.”
The New Egg 2 had not yet entered mass production. The units displayed at the Coppermine event had been every last one they could scrape together. That was inevitable, since the HPC Class-certified memory, 3D accelerator cards, sound cards, and other components installed in the New Egg 2 had actually been produced in small quantities in the manufacturers’ laboratories. Even with Coppermine technology, changing part of the semiconductor process was unavoidable. The three companies that had been contacted first had already begun converting their processes, but mass production would still take at least several more months. The time required for semiconductor factories to retool might even exceed the time it had taken to develop the Coppermine technology itself.
Fortunately, Mirae Electronics’ second semiconductor plant—commonly called the Daejeon plant—had already completed preparations. The biggest remaining issue was replacing the wafer cleaning solution to match the Coppermine process, but since the plant had only just been completed, no chemicals had been introduced yet. That allowed them to add the Coppermine process immediately and begin churning out HPC-grade memory right away.
“Besides, the most important piece—the HPC Class CPU—still needs to arrive before we can call it finished. If we rush it out just because the response is good, we could damage the Egg PC’s identity.”
More than half of a computer’s performance was determined by the CPU. The CPU installed in the New Egg 2 shown at the event had been a standard Intel Pentium. It had been a half-finished system, severely unbalanced. The final puzzle piece for the New Egg 2 was an HPC Class-certified CPU. Whether from Intel or AMD, the model that adopted the Coppermine process fastest and delivered the strongest performance would be mounted in the New Egg 2.
—Whew, if you say so, there’s nothing I can do. Lee Yong-kwon was deeply disappointed that an early launch of the New Egg 2 had slipped away. Yet he still had more to discuss.
—By the way, how’s the terminal coming along?
The matter Lee Yong-kwon was personally handling alongside the Egg PC was the launch of TG Mobile. Contracts for repeaters and management servers had already been finalized, and they were now purchasing sites for repeaters to ensure effective nationwide coverage. Because Shinsedae Telecom had also entered the fray, fierce competition over locations had broken out. However, Shinsedae Telecom—jointly founded by POSCO and Kolon—had its management split into two factions, slowing decision-making by a beat. That gave TG Mobile, always one step ahead, the advantage. They had also secured the 011 prefix and were actively conducting pre-launch marketing.
“Of course.”
Yoo Jae-won replied briefly, as if there were no issues. In truth, he had not even begun designing a feature phone yet. Feature phones had never been on his list of priorities. Still, all the technical preparations for manufacturing one were complete. High-density substrates and other key mobile components were being produced by companies owned by the New Japan Investment Bank, and Qualcomm was developing the crucial modem chip. Moreover, ID Investment had long been one of Qualcomm’s major shareholders, and cooperation was so smooth that they were even sharing the Coppermine license with Qualcomm. At this point, it seemed there was no mobile component Yoo Jae-won had not already claimed. He had even established his own LCD factory; excavation work was finished and ground preparation was underway. The plan was to quickly erect the factory building using H-beams and prefabricated exterior panels while installing state-of-the-art internal equipment. Negotiations for an LCD technology license with Japan’s Sharp were nearly concluded—positively, of course. Through this, the first feature phones would be equipped with LCDs capable of at least text messaging and simple games like Tetris.
The last remaining piece was an embedded operating system for feature phones. If the device only needed to make and receive calls, no operating system would be required, but to implement the functions Yoo Jae-won wanted, a very small, lightweight OS was necessary. Since that task was well within ID Group’s capabilities, he spoke confidently.
“At the earliest by the end of this year, and at the latest by early next year, I’ll show you a working prototype.”
Yoo Jae-won even guaranteed the timeline.
—That soon? The mobile communications pilot service was scheduled for next summer, so Yoo Jae-won’s schedule was roughly one quarter ahead. He was not saying work would begin immediately; he was simply informing Lee Yong-kwon that he had his own timeline. After wrapping up the ID Talk conversation with Lee Yong-kwon, Yoo Jae-won briefly checked on the company’s overall situation. Because everything had been arranged so that a single glance at ID Talk revealed the full picture, there was no need to summon Kim Dae-seok.
“Hmm, Nextcomcast’s ADSL adoption rate is a bit slow. That’s the only flaw.”
That was Yoo Jae-won’s single comment after reviewing the overall situation. ADSL had officially launched in Korea through Nextcomcast and was now accepting reservations in the United States as well. The problem, from Yoo Jae-won’s perspective, was that the number of applicants was not particularly high. Korea was still in the era when 486 machines dominated, so there simply was not much one could do on the internet. The U.S. had decent latent demand, but the backbone network itself was not yet complete, making service impossible for the time being. Optical cables for the backbone were being laid diligently, and the Clinton administration had passed legislation allowing cable television networks to be used for internet and telecommunications services. Once the tedious and expensive backbone construction was finished, service could begin nationwide. The free email service he had discussed with President Samuel would launch alongside it, and personal homepages would also be provided, marking the true beginning of the internet business.
“Then people claiming they’ll build search engines will start appearing too, won’t they?”
The portal already boasted Nextcom’s overwhelming market share. News services, data archives, chat rooms, and bulletin boards covering countless topics—internet usage had become almost synonymous with using Nextcom. Yet even now, new websites continued to appear. Like a small snowball rolling down a hill and growing larger, their numbers were increasing at a steep rate. Demand for a service that could search across all these proliferating websites was therefore growing as well. Under these circumstances, it would not be strange if someone came up with a brilliant idea and launched a search site. Yoo Jae-won hoped that someone would appear sooner rather than later. The burden of single-handedly pioneering the vast new continent called the internet weighed heavily on him. Of course, that burden was underpinned by confidence—the confidence that no latecomer, no matter how aggressive their pursuit, would ever overtake him.
“Third investment product sales are also proceeding as expected.”
That remark followed Vincent Greenhill’s report from ID Investment. ID Investment’s third investment product, which Korean investors and some global players had been eagerly awaiting, had finally opened. However, its structure differed slightly from previous offerings. The first and second products had fixed investment periods after which automatic liquidation occurred. The third product had no fixed term and allowed free deposits and withdrawals. Furthermore, the investment focus had been disclosed in advance: concentrated investment in the IT industry. Some investors who lacked faith in IT would withdraw, but many others welcomed the clarity. Especially in Korea, ID Investment branches were flooded with people carrying cash, to the point that the phrase “blind investment” had begun circulating. The phenomenon was so pronounced that articles appeared suggesting the government needed to intervene. Those articles were undoubtedly commissioned by rival investment banks green with envy at ID Investment’s success. Yoo Jae-won did not view the bandwagon effect negatively. Even if people were simply following the crowd, more individuals would get to taste the fruits of America’s coming IT boom. The more such wealthy people there were, the more buyers there would be for the expensive products ID Group would release.
“By the way, why hasn’t Ilsung said anything yet?”
The thought of Ilsung suddenly surfaced because of Intel’s final Coppermine license contract that had just arrived after ID Investment’s report. At the HPC Expo, there had been an Intel development team leader who had been anxiously waiting. True to his promise, the very next day Intel’s CEO, Andy Grove, had requested a meeting. Yoo Jae-won had responded immediately, and the appointment was set right away. Their relationship had consisted solely of exchanging a few letters. There had also been minor friction the previous year when Intel’s performance had dropped sharply due to security patches. Nevertheless, their shared identity as engineers and developers allowed their conversation to flow smoothly. Naturally, the license agreement had been successful. Ilsung Electronics, on the other hand, remained completely silent. Yoo Jae-won held considerable affection for Ilsung Electronics. Yet setting personal feelings aside and approaching the matter purely from a business standpoint, he had still submitted a proposal for the Coppermine process to Ilsung Electronics. After all, Ilsung Electronics currently held a seven-percent share of the global memory semiconductor market. Whether the delay stemmed from the slow decision-making typical of a massive conglomerate or from the private grudges of its chaebol owner remained unclear.
Ding!
Just as he was pondering Ilsung, Yoo Jae-won’s ID Talk chimed. “Huh? High Tech? And it’s from Redmond High Tech too? What could those guys possibly want?”
Surprisingly, the alert had come directly from the Redmond High Tech Research Institute, which contacted him perhaps once a month at most. Opening it revealed a short message—“We did it!”—along with an attached ten-second video file.
“What exactly did they do?”
Yoo Jae-won immediately began the download. The moment it finished, he hit play—and his mouth fell open. In the footage, apparently shot inside a warehouse, a flying object was hovering in midair. Startled, Yoo Jae-won replayed the video. On the monitor was a quadcopter drone with four propellers, crude but unmistakably functional.