ANTenna Blog -- How-To

A recent Gartner report says that the growth of the server blade market is likely to be slowed by a lack of hardware standardization that locks in users. I beg to differ.

Gartner, one of the leading market research firms in the high-tech field, recently published a report on the blade server market predicting healthy growth, but implying that the market's growth would be faster where it not for the proprietary nature of each vendors' server blades, which lock the buyer into a specific server vendor.

Yes, they are proprietary. But the lesson offered by the death of the minicomputer industry two decades ago is that lock-in does not come from proprietary hardware, but from a proprietary operating system. The minicomputer vendors, who each had a proprietary operating system, saw their sales evaporate as their users managed to port their applications to MS-DOS, and ever after had their pick of inexpensive PC hardware.

Today, a user with an application running under, for instance, Microsoft Windows Server 2003, is not locked in as long as there is a selection of servers running that OS.

Meanwhile, if each of those servers has proprietary features that differentiate it from the competition, so be it—the vendors want that differentiation in order to compete, and the users ultimately benefit from a competitive market. If the vendors adopted some sort of USB-like interface so that all blades could fit into all chasses, there could only be competition based on price, and the development of new features would cease.

The Gartner report went on to make other predictions that seem more agreeable, such as server aggregation whereby two blades can become one virtual server; faster I/O; more flexible I/O control; embedded hypervisors for native virtualization; and RAM aggregation in order to scale workloads.

Go to the bMighty Server How-To Center