Marvin - re: I also question the sponsor of the study, and timing. Why release 1999 data now?
There is no "Sponsor" for IDC studies - IDC makes money from yokels like me who pay for access to their data. As I said earlier, it is not IDC who said that IBM had outsold SUNW, it was IBM - IDC just puts up the numbers and leaves it up to subscribers to understand and interpret, for the most part...
As far as why they ar releasing 1999 data now, the 4Q numbers are just coming in for the more arcane categories. This is done by triangulating between sales info provided by manufacturers (which probably was available in late January and early February) with independent surveys of the distribution channels and customer base, which take a few weeks to accomplish. Finally, IDC restricts the use that subscribers can make of the data, so even though subscribers may have had some or all of this information a few weeks ago, they would not have been able to publish anything which referenced the data. If IDC did not restrict the publication of their data in this way, they couldn't stay in business. The reason people pay big bucks for access to their analysis is because they get the information edge. If subscribers could publish immediately, then a few subscribers would put the info out and people like me would not have to pay - after all, I can wait a few days, but a few weeks might make a difference.
Without a detailed understanding of what the IDC numbers really mean, the IBM press is pretty meaningless. This is like the "workstation" category a few years ago - CPQ and DELL introduced X86 based products which could do some low end workstation jobs. DELL in particular took the high end of their PC desktop line and re-labeled them workstations. So all of a sudden the X86 workstation market was exploding - at least if you looked at DELL's PR. In reality all that happened was that a bunch of PC desktops were being compared to RISC workstations, even though there was no real change in what people were buying - it was apples and baseballs. It was not until the app vendors started to write for the X86 platforms that there was any real shift in workstation buying patterns. But of course that was the point of the PR... to get the ISVs to see a market which would induce them to port.
The processor debate is also hollow. SUNW has never been the leader in pure processor horsepower - in fact SPARC has often been at the back of the pack. None the less, the overall architecture does a fine job of providing a wide range of products with great headroom. the fact that a 4-way IBM box can outperform an 8-way UE3500 doesn't mean much if there's no way to integrate the IBM 4-way with the rest of the IBM line.
As far as reliability goes, IBM has a good story, and SUNW only an adequate one. In my mind, this is an area that SUNW needs to address more aggressively - they need average RAS numbers more like AS400 and S390, not just a few sites which have those numbers. |
