The Price of Performance
It would have been fortunate if Fantasy Universe had collapsed on its own. The real problem was that while Intel, which had pushed Rambus DRAM, suffered only minor losses, memory semiconductor companies like Micron, Infineon, and Toshiba—who had trusted Intel and backed Rambus DRAM—took damage severe enough to shake their foundations. Rambus DRAM offered strong performance, but it carried a fatal flaw: price. Its exorbitant cost nullified every advantage. Furthermore, Rambus DRAM could not be installed singly; it required at least dual-channel configuration, and quad-channel for maximum performance.
“Hmm, how much is a 512-megabyte DRAM module right now?”
“One hundred twenty dollars,” Kevin Johnson replied. Yoo Jae-won let out a low whistle. It was even more expensive than he had expected. After all, the standard memory specification for PCs these days was one gigabyte. Originally it would have been around 512 megabytes, but because ID Electronics was churning out memory chips every day, the memory capacity installed in ordinary PCs had increased dramatically. Filling one gigabyte cost two hundred forty dollars. The price still drew complaints, yet it was far lower than in the past. There was one exception, however.
“And Rambus DRAM?”
“Four hundred forty dollars.”
“For one gigabyte?”
“That is the price for a single 512-megabyte module.”
“So you are saying it would cost eight hundred eighty dollars just to reach one gigabyte. And Intel made this the default choice?”
As Kevin Johnson had said, Rambus DRAM cost four times as much as conventional memory. Even Yoo Jae-won found the figure burdensome. Though his personal wealth was enormous, he still maintained clear standards for what computer components should reasonably cost. Having assembled his own machines rather than buying pre-built corporate models in his previous life, he knew exactly where the line lay. In that sense, the iWorks and NewEgg series sat at the very top of the premium tier he had defined—powerful, refined, and unmatched. Even the most powerful iWorks workstation could not justify adopting Rambus DRAM; eight hundred eighty dollars for a single gigabyte was simply too much.
An iWorks typically came with eight memory slots. Installing four one-gigabyte modules produced four gigabytes total. Switching to Intel’s next-generation platform would therefore mean spending three thousand five hundred twenty dollars on memory alone.
“Is the performance difference that large?”
“Numerically, yes, it is quite significant.”
Kevin Johnson immediately displayed benchmark data comparing Intel’s new CPU paired with Rambus DRAM against AMD’s new CPU paired with DDR memory. In read and write throughput, Intel led AMD by roughly fifty percent. In most games and office applications such as ID Office, however, the gap narrowed considerably. Even with faster memory, the CPU had closed much of the difference.
“Intel is aware of the memory-price concern,” Kevin Johnson continued. “They expect that once the platform is standardized, memory manufacturers will ramp up production and prices will fall sharply. They were hoping ID Electronics would participate in that ramp-up.”
Yoo Jae-won gave a wry smile. Releasing a new product was a company’s prerogative, but survival in the market ultimately depended on consumer choice—and even he found predicting that choice difficult. Although ID Group appeared to lead the market, every product it released had been meticulously researched using knowledge from his previous life. That was why he remained deeply skeptical about mass-producing Rambus DRAM. The issue was not licensing fees; Rambus Inc., which held numerous patents on high-speed memory technologies including Rambus DRAM, was already majority-owned by Yoo Jae-won. The notorious licensing costs of the past no longer applied. Even so, the high price stemmed from technical complexity and low-volume production. If Intel’s rosy projections came true and large-scale manufacturing began, prices might drop enough for Rambus DRAM to succeed after all.
“Hmm.”
Though Intel’s argument held some appeal, Yoo Jae-won did not change his mind. The memory technology chosen in his master plan was DDR. DDR already had a roadmap extending far into the future, whereas Rambus DRAM had nothing prepared for it.
“If Intel makes a formal request, we must clearly decline.”
“Understood.”
Kevin Johnson accepted the order without hesitation. As president of Android, the choice of memory standard made little difference to him.
The next day, as Kevin Johnson had predicted, Intel formally requested ID Electronics’ cooperation in producing Rambus DRAM. Yoo Jae-won instructed that a reply be sent exactly as he had decided: production volume could be increased modestly, but a full-scale process conversion was impossible. He also attached the suggestion that Intel focus on DDR instead. Although DDR currently lagged behind Rambus DRAM, raising its operating speed would quickly close the performance gap. Crucially, this reply was not sent in secret; it was also released as an official press statement so that everyone could see it. After all, the types and quantities of products leaving the factory could not be hidden. If anyone wished to know, they would find out anyway. Issuing a public warning to other companies seemed prudent, and it might also reduce the scale of Intel’s misstep. For the next twenty years, ID Group would still need to source CPUs from both Intel and AMD. Any blunder by Intel ultimately harmed the entire market. Fair, sustained competition between the two companies benefited both consumers and businesses. Yoo Jae-won therefore hoped Intel would heed his warning.
Unfortunately, Intel’s response did not arrive as quickly as expected. After reviewing the development status of Android’s next operating system, Yoo Jae-won, who had been in Seattle, returned to San Francisco on October 31.
October 31, 2001. On the calendar it was an ordinary day. For gamers—especially those who had eagerly followed Warcraft—it was a date they had counted down to for months. The open beta of World of Warcraft, whose development had been announced long ago, was finally beginning.
An open beta meant anyone who wished to try the game could do so without special qualifications. It was the moment when the game’s overall quality, its alignment with player tastes, and—most critically—its server stability under an unprecedented flood of simultaneous users would receive a comprehensive public evaluation. Blizzard Entertainment had therefore made every preparation with assistance from its parent company, ID Entertainment. The highest priority had been server capacity. Blizzard had secured the maximum computing power it could muster. Judging even that might prove insufficient, it had also installed the World of Warcraft server software on ID Technology’s cloud servers so they could be brought online at any moment.
In Korea, it was still common for companies to boast that their servers had crashed on launch day because too many players had connected at once. To Yoo Jae-won, who considered himself a leader of the internet era, such a situation was nothing short of disgraceful. Whether one million or two million players connected simultaneously, every gamer should be able to enjoy the game without inconvenience. That was what mattered to him. Blizzard Entertainment fully agreed. Even if the number of players dropped sharply after the open beta ended, the servers would remain at maximum capacity for the duration.
“I’m so jealous,” Tiffany said.
A grand opening ceremony celebrating the World of Warcraft open beta was scheduled, and Yoo Jae-won was expected to attend and lend his presence. Designed as a festive party rather than a formal event, it aimed to excite gamers. Yoo Jae-won was busy preparing for the party while Tiffany was getting ready for work, so her envious remark came naturally.
“You should leave work early today and come straight to the venue!”
Yoo Jae-won’s suggestion was easy to make, but the current atmosphere inside Chevron made it difficult to act on. In the unprecedented situation following Jacob’s death, Tiffany and the other younger contenders were locked in fierce competition under Frederick’s approval. Frederick had not simply summoned his children; he had assigned each of them specific tasks. Tiffany’s assignment was the merger with Texaco—the very project Jacob had been leading. The symbolism was unmistakable. While Jacob had been tasked with directing the merger, Tiffany was assigned to assist Frederick. Because Texaco was itself a giant oil company with a market capitalization exceeding two hundred billion dollars, Tiffany could not handle it alone. Frederick therefore took the lead while Tiffany supported him.
Tiffany’s aunts had reportedly protested the news loudly. The project had been Jacob’s. Even with Frederick at the helm, its symbolic weight remained unique. Yet after their disastrous outburst at the funeral, they dared not voice open anger. Although Tiffany currently led the succession race, she too felt considerable stress. She had even asked Yoo Jae-won for advice on managing much older subordinates. Unfortunately, he had been unable to offer useful guidance; their situations were entirely different. Yoo Jae-won had built ID Group from nothing, hand-picking every founding member. People who dismissed him for his youth had never been allowed inside the company in the first place. Tiffany, by contrast, had entered Chevron—a company with nearly a century of history. Adapting to a pre-existing organization was no easy task.
“I’ll keep your seat warm for you.”
For now, all Yoo Jae-won could do was offer wholehearted support. Of course, preparations to provide decisive assistance at the critical moment were already underway. In any company, the ultimate source of power was stock ownership. No matter how capable an executive might be, they could never override the rights of shareholders. That was why Yoo Jae-won was steadily accumulating Chevron shares.
After seeing Tiffany off, Yoo Jae-won returned to the living room and began tidying up before heading to the office himself. He was about to turn off the television when a breaking news report appeared.
In the past, the Taliban would have joined forces with al-Qaeda to resist U.S. forces, but the successful separation of the Taliban and al-Qaeda had prevented the conflict from expanding. Moreover, Osama bin Laden’s hidden charisma as al-Qaeda’s leader was rapidly crumbling. Evidence recovered by Delta Force from one of his safe houses was being released piece by piece, and much of it was difficult to reconcile with the image of an Islamic extremist. Alcohol was the least of it; his computer had been filled with adult video files. The list had quickly leaked online. Although such material should never have been made public, U.S. intelligence had orchestrated the leak to prevent bin Laden from becoming a martyred hero.
Yoo Jae-won paused again. If Afghanistan was being resolved, then China was only just beginning. China’s handling of the cracker group Skidrow was exactly the tail-cutting exercise he had anticipated. Rooting them out completely would have required investigating even the highest levels of the Communist Party—an unacceptable risk for President Hu Jintao. If even one of Skidrow’s patrons turned out to be one of his own people, the damage would be catastrophic. Naturally, the United States reacted with fury. No full investigation of the Skidrow members had been conducted, and China had refused both U.S. observers at interrogations and any extradition. Having declared there would be no compromise regarding the 9/11 attacks, America’s next step was an inevitable head-on collision with China.