Typical. You wait ages for a big IT announcement, and then two come along at once. Last week, not only did AMD officially unveil its long-awaited 64bit Opteron chips, but Microsoft officially launched Windows Server 2003.
The AMD launch was preceded by Microsoft's confirmation that it will support the Opteron with a 64bit version of Windows Server 2003, and will also launch a version of XP to support AMD's Athlon 64 desktop chip, due later this year.
This news probably caused a certain amount of displeasure at Microsoft's partner Intel, but you have to see it from Microsoft's point of view. AMD's technology has built up a great deal of interest around 64bit PCs, and it seems that many such systems will be bought by early adopters in the coming months.
Microsoft would, of course, rather see these systems ship with Windows pre-installed, so an x86-64 version of its software was almost a certainty.
AMD has been showing demo Opteron systems with a beta of 64bit Windows for some time, so it was really just a question of when Microsoft would come off the fence and risk incurring Intel's wrath.
Intel and Microsoft are of course the two behemoths that form the so-called Wintel alliance that dominates the computer world with the combination of Microsoft operating systems running on Intel-based hardware.
But Intel has not been shy in wooing other partners. There's HP, for example, which co-designed the Itanium architecture. However, HP is much more interested in pushing Linux and HP-UX than 64bit Windows; and Linux on Itanium also features highly in IBM's plans. Microsoft probably isn't too happy with this, as it sees Linux as the major threat to Windows.
These developments might point to a break-up of the old alliance, but both Microsoft and Intel are probably just spreading their bets to cover all eventualities in these uncertain times. There is enough room in the industry for more than one 64bit architecture and more than one 64bit operating system.
AMD doesn't have much traction in the enterprise market, so its strategy is to target small-to-medium-sized businesses first, and work upwards.
The promise of compatibility with current 32bit software and a painless transition to 64bit software later ought to work like a charm in this market, and if enough smaller firms invest in Opteron systems, AMD might gain the momentum needed to draw in the more conservative blue-chip firms.
But it is always foolish to discount Intel. The chip giant is playing a longer game with the Itanium architecture, which it expects to last for a good 20 years or so. Itanium systems right now are overpriced and do not meet the hyped-up performance promises, but Intel has a reputation for delivering the goods, and it is quite possible that the Itanium family might eventually be as ubiquitous in the datacentre as Pentiums are on the desktop.
My guess is that AMD has a good chance of breaking into the enterprise market in the medium term. Most large firms are in no mood to invest in new technology in the short term, and the Opteron has yet to prove itself anyway. In the long term, the smart money will always be on Intel. Unless Sun Microsystems has an ace up its sleeve, that is...
reader comments