Intel’s dual-core delay really hurts

When Intel announced last week that the delivery of its dual-core Itanium would be delayed by three months, from the first quarter of 2006 to the middle of the second quarter, even the keenest ear could not discern the merest peep of disappointment. But if you listen just right, you can hear the chirping of the cell phones carried by AMD, IBM, and Sun Microsystems reps worldwide.

The whole of the computing world isn’t perched on the edge of its collective seat in anticipation of dual-core Itanium. But in the science, technology, supercomputing, and massive enterprise markets where pockets are bottomless, most applications are custom-written to standards, and the appetite for computing performance and throughput is insatiable. Every six- or seven-figure purchase order is an opportunity to identify and invest in the new leading edge; in many cases, getting the most firepower for the dollar is a purchasing imperative. A one-quarter delay for dual-core Itanium means that for three months, Intel and its Itanium OEMs won’t be at the table for high-performance computing purchases.

Don’t sneeze at that: Unlike most of IT, buying cycles in high-performance computing are driven as much by the debut of new, faster technology as by need. To snag that high-dollar, high-profile business, a vendor has to bring exciting new gear to the table. Not only will Intel miss out on performance-driven buys for three months, it also suffers the ignominy of having its flagship microprocessor arrive last, and, with the delay — dead last with a bullet — to multicore. Direct and indirect losses from Intel’s delay will tally in the millions of dollars, not counting continuing revenue from services, upgrades, and support. And there’s no way to count the OEMs and end customers that will just give up waiting because it seems like Intel is less and less serious about making Itanium a real contender.

The setback is the latest in a string of events that raise questions about Intel’s ability to execute on its innovative R&D, a commodity of which Intel undoubtedly has plenty but which Intel hasn’t been able to get to market. There has always been a tug-of-war at Intel between x86 and Itanium over budget, staff resources, and scheduling priorities. Time and time again, Itanium loses out to the can’t-miss x86. Intel’s latest focus on low-end and power-efficient systems, including digital entertainment technology, reflects a campaign to carve out new volume niches that competitor AMD can’t quickly attack. Even if the dual-core Itanium delay has nothing to do with these other factors, Itanium is certainly a poor fit with those elements of Intel’s new strategy that we’ve seen to date.

Although analysts will debate the relevance of the one-quarter delay in the release of dual-core Itanium as an isolated event, this latest setback prolongs Intel’s long-awaited return to form.