Three Minutes With Patrick Gelsinger

Patrick Gelsinger is vice president and general manager of Intel's Desktop Products Group. He joined Intel in 1979, and worked on the i386 and i286 chip-design teams. Gelsinger spoke at the recent Intel Developer's Forum and during a break, PC World asked him about the company's rocky 1999, and its promising 2000.

PCW: Intel hit snags last year with the 820 delay and reported Pentium III shortages. Did that hurt the company?

GELSINGER: We certainly were hurt by that, and our customer relationships were hurt by it. We had [PC vendors] who basically were forced to bring competitive products to market because they couldn't satisfy customer demand. Part of the problem is we took on a great deal last year and we really took risk-taking too far. I think collectively we over-reached in 1999, and we are going to fix that in 2000.

PCW: Analysts say AMD's Athlon success is good for consumers and the industry. Is it good for Intel, too?

GELSINGER: Competition is always good. Competition forces you to work harder, run faster. How have we responded to competition for the last 20 years? We've done three things: superior products and technology, superior manufacturing and execution, and superior branding and marketing. And that's how we're going to respond in 2000.

PCW: Talk about Intel's lineup for 2000.

GELSINGER: We have an awesome product line this year. We're going to have Itanium, this whole new architecture at the high end, to compliment Xeon. We have a new microarchitecture called Willamette that fundamentally changes the rules of the game. And [at IDF] we showed production-level hardware on HP, IBM, and Dell at 1 gigahertz, shipping into the market shortly. And at the low end is the first ever product [code-named Timna] that is architected uniquely for the value segment.

PCW: You recently demonstrated a 1.5-gigahertz processor, but who needs that much power?

GELSINGER: On the consumer side it will enable awesome Internet. All of a sudden you have enough computing to be able to do some things that were just unrealizable before. Where you have multiparticipant 3D worlds, you can move from representing [images] geometrically to parametrically, and you literally go from multiple megabytes, to tens of kilobytes of information. You can't shove that over a 56 kilobits-per-second connection, so it's going to enable entire new user experiences. [Note: Parametric images require substantial processing power but little bandwidth, while geometric data takes more bandwidth and less processing power.]

PCW: Let's talk about Timna. Where does it fit into the product line?

GELSINGER: Doing the coolest, fastest, baddest machine, that sort of becomes a genetic makeup for us. Typically we do the coolest thing we can, and then migrate it down the product line over time. With Timna we offer the most cost-effective solution that we know how to do for the value end of the marketplace. So we're going to launch it at the bottom and bring it up.

PCW: At IDF you restated Intel's support of Rambus memory, despite its rough start. Why is Rambus important?

GELSINGER: Let me answer a bigger question for you on Rambus. What's different about Rambus in 2000 versus 1999? In 1999 it didn't work, it wasn't available, and so you couldn't demonstrate real value for it. Other than that, it was great. What we showed yesterday was value, it works, and improving availability. We're going to ship multiple millions of 820 platforms this quarter, and customers are seeing real demand for their products. Rambus is the technology for the desktop PC. As you get up to 800 megahertz and 1.5 gigahertz you get to 10, 15, 30 percent [improvements] on today's application benchmarks. That's a huge number. You're starting to see meaningful benefit on some applications.

PCW: Which products in 2000 are you personally most excited about?

GELSINGER: I've been on this style kick for the last year and a half. Ease of use and style. PCs are too hard to use and they break too easily. What we are trying to do is create a whole new level of excitement in the industry. We've seen the first products: [Dell's] Web PC, [Compaq's] Ipaq, and [Gateway's] Astro. But we think with new products and technologies such as Windows Me [Millennium Edition], instantly available, USB [Universal Serial Bus] 2.0, and the small form factor things that we've demonstrated here, PCs are just going to be really cool fashion statements next year. That certainly gets me out of bed in the morning.