day 11

the next big thing 3: fry pilots

So far this week we've visited Fry's, the fabled computer supermarket, and we've met Buck the engineer, a typical Fry's customer. Everyone, all 5 billion of everyone, wants to shop at Fry's, to be a Fry Pilot. This is the game plan. The game is called "stoking the global star-making machinery." It consists of a lot of engineering and even more Hollywood. Eventually everyone will get the chance to be a Fry Pilot. But first, they need a low-cost Box. They need a low-cost Wire. They need low-maintenance, pay-as-you-go software. And most importantly, they need to be compelled. Just about everyone in Silicon Valley knows this, but no one is sure how to pull it off. Remember: Gordon Moore declared that the processing power of microprocessors will double every 18 months if cost is held constant. This law has propelled the industry into orbit through a series of ever-increasing rates of obsolescence. Whether we need it or not, the Pentium Pro is being crammed down our throats, and the excuse is that the machine we bought last year is now out-of-date. Intel is working overtime to rid itself of the Pentium chip so it can sell Pentium Pros to rabid consumers who think they need faster and faster machines. But wait; Moore's Second Law says that each generation of chip-making factory costs twice as much as the previous one. The number of chips that must be sold every 18 months to pay for these multi-billion-dollar facilities gets larger and larger. Add to this the fact that capitalists need to get at least 15% return on their investment or else they lend their money to build gambling casinos and oil refineries instead. Soon the bulk of chips that must be manufactured and sold becomes an impracticality, a bridge too far, an impossible dream. So Buck and his fellow engineers must invert Moore's first law. Instead of building faster and faster chips, Buck must now build cheaper and cheaper systems. By holding the performance at a level of about 250MHz (the original Pentium was 60MHz), the cost can be reduced, instead of boosting performance. Investment now shifts to the care and feeding of a declining price curve. How will Buck do this? Can Fry's be saved? Will the Next Big Thing be a Box with 250MHz processing power? How will Silicon Valley dominate the world for the next decade or two?

the next big thing: 1 2 3 4 5 Daily Dose Index