In its latest "Hype Cycle" report on Linux Gartner Group has suggested that the next target for Linux applications should be the data center. (Just to prove I can snark, too, that's the Columbia University data center in the 1960s at right.)
Gartner found that for eight processors and beyond, Linux must demonstrate performance, security and application proof points based on the 2.6 version of the kernel and that the biggest test continues to be whether it can function as a data centre server for mission-critical applications.
Given that IBM is the unquestioned leader in data centers -- Gartner specifically mentions its Websphere application as a center of activity -- you can ask how much of a challenge this is. Much of the pace of change here is really up to IBM, which continues to sell its proprietary Unix, AIX, into data centers.
Beyond this, Gartner has a loud warning for Linux advocates, and Linux skeptics, namely don't believe the hype:
Mainstream business use of Linux is nearing the Peak of Inflated Expectations, where the costs of migration may exceed the cost benefits. This phase is characterised by over-enthusiasm and unrealistic projections when flurries of well publicised activity by technology leaders results in some successes, but more failures, as the technology is pushed to its limits.
All this would be much more interesting to me if Gartner's report indicated any understanding that Linux is a movement based on Unix, not a proprietary technology driven exclusively by IBM or anyone else.
But your Fortune 500 CIOs now have their Talking Points. It's up to Linux advocates to meet them with performance.