Mike -
Looking at Dell servers, it seems that they've got redundancy in the areas you mention as well as support in software for failure recognition. Do you feel that these are not as reliable as the others you listed?
Dell's server strategy is being driven by Mike Lambert, who used to be VP of Marketing for CPQ's systems division in 1994-95. He learned a lot about CPQ's server design philosophy and strategy from Gary Stimac, CPQ's legendary employee #4, who was responsible for the creation not only of CPQ's leadership position with the 1986 release of the Compaq 386, but also CPQ's server designs in 1989 which created the PC server market.
Mike Lambert is executing CPQ's strategy from 1994, and doing it well. He has a big edge in that CPQ has plowed the ground, and he has the additional advantage of Dell's great field execution, something he could never count on at CPQ. Dell's server products are good and getting better.
But the key to getting to better than 99% reliability is not just hardware, but also careful management of actual installations in the field, and advanced engineering work with the partners, particularly MSFT.
I'll give you a concrete example. Looking at the MSFT web site, it appears that CPQ has about 175 certified cluster configurations on the MSFT hardware configuration list. Next best is IBM with 15, HP with 14. Dell has 5 configurations certified. That reflects a level of engineering commitment that Dell probably does not want to make, since their focus is more on mfg execution.
Each configuration takes trained field people to install, configure, and maintain. Compaq, IBM, HP, and DG have all invested in the field people to make that happen, Dell has not. DG has done less in development than the other 3, but makes up for that with a very aggressive high end services component. CPQ is obviously the leader in that space, with more than 10 times the supported configurations, and more than 10 times the trained field staff of the next best.
Look at the PCI-X specification. CPQ. IBM and HP worked together to develop a high end bus specification. Why? Because those three have enough market presence to define a standard, and the engineering capability to make it work. Dell may use it but they could not contribute to the creation of something like PCI-X.
So Dell has hardware designs which could compete in that space, although they are not at the level of CPQ, IBM and HP. But they can not put together the complete offering. My sense is that Dell will take the low hanging fruit in the enterprise space and continue to focus on sales and manufacturing execution as their differentiators.
Thanks for the options info. I was more interested in your philosophy, so that's the response I was looking for. |
