Exclusive: The Xbox 360 Uncloaked - Page 6
- May 05, 2006 16:42 PM PST
- Email this!
Microsoft wasn't going to copy Sony and create a full custom chip with its own factory. It didn't have the time or the inclination. Microsoft had to tailor the system to run video games, without going full custom in a risky project that might take years to finish. With the original Xbox, Microsoft had chosen mostly off-the-shelf parts developed for the PC. That saved a lot of time. But it meant that the Xbox life-span wasn't as long as a box that had custom-designed silicon. Other off-the-shelf parts would come along and quickly enable the PC to move ahead of the game console. Microsoft had more time to do custom parts this time, but still wanted to take advantage of PC hardware R&D. It also wanted to take advantage of the vast growth in contract manufacturing infrastructure around the world. It could use somebody else's factory instead of its own and use that to beat Sony, a cash-strapped company trying to fight off competitors on too many fronts.
Todd Holmdahl, the head of hardware for the Xbox, knew where he wanted to go. He had known how to make the mouse business work without owning factories. Microsoft hired contract manufacturers such as Flextronics to make its mice and keyboards, and it did the same for the original Xbox. Running the business in a virtual way was natural.
Holmdahl had known about Nintendo's deal with graphics chip maker ATI Technologies on the GameCube console, and he also considered WebTV's experience in system design. ATI designed a graphics chip for Nintendo, but it allowed Nintendo to take the design to a contract chip manufacturer and have it fabricated on its own timetable. ATI generated profits on engineering fees and royalty payments for every box sold, but Nintendo owned the designs and didn't have to pay ATI a margin for making the chips. Nintendo owned the intellectual property that ATI had created, and that mattered a lot in terms of profits. Holmdahl wanted the same for the next Xbox.
"The first time, we had a short period of time and it didn't afford us the opportunity to make these relationships," Holmdahl said. "The whole intellectual property business, sharing IP, was just in its infancy. The lesson we learned is you want to control your destiny." 1
Holmdahl knew that the rush to get to market had forced Microsoft to sign some contracts that it eventually came to regret. The company tried to lean on suppliers to cut costs as much as possible, but Nvidia's CEO, Jen-Hsun Huang, had fought back. The battle spilled into public. Both companies sought a resolution in an arbitration process, which was the only thing that prevented them from going to court.
Huang wanted to redo the contract because the yields on his chips weren't as high as expected. Nvidia had to scale back the speed of its graphics chips to get better yields, but the problem persisted. Huang threatened to stop shipping chips. Microsoft objected, and the arbitrator sided with Microsoft.
The lesson inside Microsoft was clear. Holmdahl wanted more control over the design and manufacturing of its chips. That would give him more leverage over suppliers. This was the key for Microsoft to control the costs of the box. From Microsoft's point of view, Nvidia and Intel had kept too much of the control in terms of cost-reducing their chips and determining the price changes on their own schedule. Because they didn't cut the costs quickly, Microsoft could not cut the prices on the original Xbox. Yet, for the sake of compatibility, it seemed Microsoft was stuck with those vendors. It was almost a given that old Xbox games had to play on Xenon.
The last thing Holmdahl needed was a repeat of the PC story. In 1981, IBM was in charge of the design of the system for the personal computer, but in its rush to market it unwittingly gave away most of the technical decision-making - and profits - to Intel and Microsoft. IBM eventually exited the PC business, while Intel and Microsoft became dominant firms in PC technology.
