Linux on the desktop has been the subject of much discourse and debate of late. Recent announcements suggest perhaps it's part of the strategy of companies like IBM and Sun: Keep everyone focused on the desktop while they quietly take over the server.
Although IBM and Sun are founding and primary members of the ABM (Anyone But Microsoft) crowd, their tactics on the server front are at odds. IBM began its Linux offering on high-end servers while Sun offers Linux on the low end, saving the high-end market for its Solaris OS.
Sun needs to reconsider this. Long ago, Big Blue became the corporate standard because it was able to hook customers early and migrate them to high-end hardware. By offering native Linux on high-end IBM hardware, IBM has provided a convenient migration path for midmarket organizations. Sun, on the other hand, offers very few incentives to buy into its low-end hardware if Linux is in the mix. Expertise on Sun's low-end systems may not carry over into the vendor's more powerful machines.
Both Dell and IBM have dropped support for Linux on the desktop, citing a lack of demand. But the only Linux desktops Dell and IBM offered were the expensive, high-end models. These systems were unlikely to be purchased for home use and were too expensive for corporations to consider on a large scale. Sun is repeating this mistake.
Linux has made huge leaps on the server and will continue to do so. But there are plenty of organizations where it has yet to make even a small impact. It is unlikely that an experimental Linux server will be installed on expensive, high-end hardware there. They'll start on the cheap boxes, and only when it's proved itself to be capable of supporting enterprise class, mission critical services will Linux be deployed on the beefy hardware.