Risc vendors have Rubicon to cross

Some have crossed it, Sun, other's ain't yet

Intel Developer Forum Paul Otellini, general manager of Intel's architecture group, was in classical mood when he hosted a question and answer session in the desert today. Otellini said that companies, such as IBM and HP, had crossed the Rubicon by wholeheartedly adopting Itanium as their future, despite having Riscy platforms. Readers with a classical bent will remember that the Rubicon is a small stream (geddit?) that separated Gaul from Italy in the Roman republic, was crossed by Julius Caesar in 49BC, so breaking the rule that generals could not lead armies out of the provinces where they were supposed to stay. Sun, however, he seemed to be suggesting, had no intention of crossing said Rubicon. He said: "A number of leading Risc-Unix vendors have made a very solid decision to move to Itanium. I'm very comfortable that some have crossed the Rubicon. For their own reasons, Sun didn't." This naturally posed the question of whether Compaq has crossed the Rubicon, given that its first Wildfire systems using the Alpha microprocessor are beginning to spawn. Otellini responded by saying that Compaq is "certainly committed" to IA-32 and IA-64, underlining that by saying it is one of the companies at the forum which is displaying Itanium boxes. But, he added, the company's decision to drop the development of NT for the Alpha platform, put the firm in an interesting position. He said that while he would never say never to the idea of Intel selling branded servers direct, his company certainly had no intention of doing so in the foreseeable future. He also talked about the introduction of Foster (Willamette) technology and its impact on the market. He said: "You'll see two and four way systems and a number of eight way systems." But the products would not compete with the IA-64 platforms. The "sweet spot" for Itanium would be four way and some eight way systems, he said. "We would like to see Itanium move early into that market. We should be able to be successful in a three year period." That market would be in the mid-range initially, because of the "compelling price-performance" features of the Itanium. Adoption of IA-64 for back end functions would be slower, he suggested. Intel was attempting to ensure that it bridged the legacy systems to what he described as the new world. ® Intel Developer Forum: Spring 2000 Full coverage

Sponsored: Designing and building an open ITOA architecture