Commodity? Not hard disk. (Storage as I see it).

I have always found it difficult to characterize hard disk drives as commodity items. To me, a commodity is 'characterized by not only ubiquity, but by slowness to change. A real commodity is the same thing, year after year after year.

Hard disk drives are not the same, year after year after year. They change constantly to meet new applications demands and processor capabilities. By way of example, I was speaking with Sam Sawyer, charged with HDD product management at Fujitsu about the next generation of mobile HDDs. Right now, most mobile disk drives top out at about 20GB on a platter. The newest 'generation from Fujitsu will chime in at 60.

How do they get there and still keep the drives affordable?

Let's face it: The current state of the disk drive industry is characterized by very tight margins. So tight that many have left the field. IBM sold their HDD business to Hitachi. Quantum sold theirs to Maxtor. Fujitsu got out of' the desktop drive business, which will end life within the next few months.

R&D is the answer and is the reason that hard disk isn't a commodity. In Fujitsu's case, the 60GB/platter introduction is a result of hard technology development and driving applications. 60GB/platter accommodates 1.2 million email messages on the average, or 60,000 high-resolution digital images. The increasing numbers of "road warriors" who take their laptops everywhere have the need for these capacities to cope with business on the run

The applications are obvious...email, digital images, spreadsheets, Internet usage and download. The metadata 'of increasingly complex applications is another consideration. All of these are enabled by smart technological innovation. Fujitsu, as this month's exemplar, has a way with GMR heads. Their specular head technology adds another layer to the GMR structure, giving the electrons some maneuvering room and improving the signal-to-noise ratio. I'm not quite sure whether this is Fujitsu's sixth- or eighth-generation head technology, but the difference is measurable as it meets and exceeds WinBench specs.

Head technology is one area for improvement, but there are others. Fujitsu's new generation of mobile drives feature improvements in the actuator assembly, as well. The actuator bearing is slightly larger, helping to stabilize the drive in a shock situation. The actuator arm features an aerodynamics-dictated change in shape, not only helping flying height but helping control operating shock characteristics as well. Firmware refinements trim the idle-to active timing down to 40 ins.

Other economies can be realized, through better production and test equipment, of course. But it is hard technology developments that make the long-term difference. The vigor of technological improvement is what convinces me that disk drive technology cannot be a commodity.

What's in Store

Fujitsu's Sawyer and I also discussed some of what's to come in the future. He told me that more and more of Fujitsu's customers are inquiring after the 2.5-inch form-factor disks, and seeing them as eventual replacements for the currently popular 3.5-inch drives used in both stand-alone servers and in automation. Sawyer pointed 'out that issues of power consumption, performance improvements, and the advantages of spreading data over more spindles draw customers. I suggested that some RAID vendors would like to get more capacity into a 1U or 2U rackmount space.

But part of the motivation might be to take advantage of the performance improvements and cost savings of an ATA-based device. Another reason that hard disk is not a commodity: One size does all. Some users and applications need all the power and throughput of a SCSI or FC-based hard disk. But others do not, and the performance delta between "low-end" and "enterprise" disks is shrinking.

Commodity? Look to another technology...not hard disk.