September 28, 2006

10,000 Itanium Apps Now Available

Earlier this week, at the Itanium Solutions Summit, the Itanium Solutions Alliance announced the availability of 10,000 applications that can run on Itanium 2-based systems. This represents 50+% growth in the number of available applications since the Alliance was formed last year. The Alliance also noted that Oracle announced that it will work with the Alliance in its certification of Oracle software on Itanium-based platforms. At the same, they quoted IDC numbers showing Itanium-based systems grew about 36% year over year (Q2 2006) and that Itanium-based platform revenue is roughly 44% POWER-based server market share, and roughly 45% of SPARC-based server market share.

When this allaince was first announced last fall, the requisite cadre of card carrying members was out and about briefing the analyst community on the Alliance’s vision of grandeur and how soon Itanium would overtake all server platforms of relevance. Roughly a year into things, Itanium has seen decent growth in the marketplace, and the alliance is more than likely one of the reasons this growth has occurred. As part of the analyst outreach last fall, there was the implicit and sometimes explicit statement the unlike other chip based alliances, read power.org, this alliance was “open” and represented the interests of many vendors, not just one. Of course, there are inaccuracies in this assessment, as power.org is about an architecture, not a chip, and it has dozens of vendors involved, not one, and it is far from “closed.” Nevertheless, after a year, we see Itanium measuring itself as having acheieve 44% of the market share held by POWER and roughly the same for SPARC. The message seems clear, we are better than those guys, we will overtake them, and the world will look like the place we want it to be. OK, perhaps, but perhaps not.

10,000 applications is a decent number, especially when considering that they are server applications. This is a far cry from the much smaller level of industry support evident just 2 or 3 years past. This is good news for organizations that have either been ordered by their systems vendor to embrace Itanium or for those genuinely looking for an alternative. While one might realize this is a fraction of the applications available for other platforms, it still demonstrates a growing commitment to the platform.

One thing that bugs me about vendors, their alliances, consortia, and other forms of industry coupling, is the urge to define oneself in terms that claim superiority to others in a very relative fashion. For example, we are 44% of POWER's marketshare, or we are growing faster than them. So, is Itanium only 44% of what it should be? This dredges the memory of 1982 Ford automobile ads where the announcer assured the audience that these new Fords were 50% higher in quality than the ones made just 2 years earlier. How did that make a 1980 Ford owner feel? Is the fact that the now 10+ years in the making Itanium has now garnered less than half of the market share of either of its main non-Intel competitors in reality all that impressive considering the number of competitive platforms that were shot by their owners who ran to Itanium to solve their market share woes? Of course, all of the focus on the “other guys” allows vendors to overlook some more pertinent issues, such as what is your identity? Comparisons are fine, but without clear definition of one’s self, it’s all about the other guy. Talk about a codependent relationship in the making.

Nevertheless, we are seeing growth in Itanium sales and the Alliance should take ample credit where it is due. But at the same time, I wish they would be a bit more honest about what they represent. No matter how often or how loudly they say it, Itanium is NOT an industry standard platform, NOR is Windows an open platform (as was claimed at the press event). Windows is a de facto industry standard given its large installed base and thousands of manufacturers providing the hardware on which it run. Itanium is a single vendor sourced chip that lacks any dejure standards bodyrecognition (other than Intel itself) nor does it have wide spread de facto deployment in the marketplace. Call the spade a spade and be truthful about it. Proprietary solutions are not inherently evil, but openly promoting one as something else is disingenuous, and one of the evils of marketing.

No comments: