For the second time this year, Intel is delaying the release of its “Tukwila” high-end Itanium processor, this time until early next year.
The chip had been expected to be released in mid-2009, according to an Intel spokesperson. Previously, the release was scheduled for early 2009, but the giant chip maker in February announced a delay.
Intel issued a statement May 21 saying that during system-level testing, engineers had “identified an opportunity to further enhance application scalability. As a result, the Tukwila processor will now ship to OEMs in Q1 2010.”
The statement further said that the change will give Tukwila a two-times performance increase over the current “Montvale” Itanium chip, which will better let it compete against RISC offerings from IBM and Sun Microsystems.
“This change is about delivering even further application scalability for mission-critical workloads,” the statement read.
However, it also means another delay in a series of delays for Intel’s high-end chip, which the company positions as an alternative to IBM’s Power architecture and Sun’s SPARC chips in the shrinking but still lucrative RISC space. According to IDC, worldwide revenue to Unix servers was $4.9 billion, representing 36.2 percent of the overall server market.
The Intel spokesperson said the Tukwila delay should have little impact on the adoption of Itanium processors. Shipments of the chip grew 10 percent between 2007 and 2008, according to IDC, and Intel officials expect that trend to continue.
However, Charles King, an analyst with Pund-IT Research, said another delay could benefit competitors.
“A major slip in the next-generation platform delivery by Intel could translate into a more competitive edge for IBM, and possibly Sun,” King said, noting that Sun could become even more aggressive after it is bought by Oracle, a company that has a strong go-to-market strategy “and knows something about stealing customers.”
IBM has had some success enticing customers to leave Hewlett-Packard and Sun for its Power architecture, he said.
The impact on the overall IT industry of the Tukwila delay will probably be minimal, particularly in comparison to the x86 market, King said. As illustrated by the release in March of Intel’s Xeon 5500 Series processors-and the accompanying fanfare by OEMs large and small-most of the interest is around x86 servers, and as rival Advanced Micro Devices found out with its “Barcelona” Opteron chip, a delay there can have a direct impact on revenue.
“The industry focus on Itanium is becoming significantly less over time,” he said. “[The Tukwila delay] is a big deal, but I’m not sure how big in the larger context. I wouldn’t think a lot of people are losing sleep over it. … The impact of Itanium in the industry is waning.”
Clay Ryder, an analyst with the Sageza Group, agreed, though he said there will be some impact felt.
“Anytime something shifts in the schedule, it’s an impact to someone,” Ryder said. “Intel is not developing these things in a vacuum. They’re making them for someone.”
Currently the only major OEM with any Itanium business is HP, which in recent months has tried to position its Itanium-based blades as viable mainframe-replacement technology and as a platform that enables businesses to keep their legacy applications, King said. In addition, there are reports that Itanium has gotten some traction in such markets as Japan, he said.
There also are other vendors within the Itanium Solutions Alliance, such as Fujitsu, NEC, Hitachi and SGI, who also are going to be impacted, Ryder said. For end users who may be looking to jump to another platform, the good news is that they have options, from increasingly powerful x86 chips to the IBM and Sun processors, he said.