Good point, Hal. I think you're on to something.
Compaq is the top global computer maker and it is slicing off layers of its fixed cost structure by adopting the 'build to order' model of Dell and Gateway. This allows Compaq to, among other things, compete more aggressively in the sub-$1,000 PC category, which in order to grow will require squeezing everybody in the supply chain to lower their prices. Compaq, for example, is using AMD and Cyrix as leverage to force mightly Intel to lower its entry level chip prices. The disk drive makers will be squeezed too if Compaq is to generate the volumes necessary to make up for the razor thin margins on its $800 PC.
In the $800 and below segment, for example, disk drives will probably consist of no more than 2 platters and at most 4 heads. If you look at Read Rite's, the leading independent head maker, road map for this year, their volume production will be in the following ranges:
TFI - 2.0 gigabytes per platter
MR - 2.1 gigabytes per platter
Heads and platters are the two most expensive components of a disk drive. KM allows somebody like Maxtor to offer some really aggressive pricing at the sweet spot capacity points.
For example, the 1.6 GB disk drive capacity point was the highest volume capacity point last year. Quantum rolled out a line of 1.6 GB gigabyte MR disk drives but they didn't produce it in volume because they couldn't compete with WDC which had a very aggressively priced line of 1.6 GB TFI drives to pump into the OEM and distributor (retail and small OEM) channels.
KM is, in practical terms, a platter saver. A platter saver is a head saver. 'More capacity at the same price' is going to be a money maker for somebody this year. A 30-40% capacity edge at nominal cost means that if your competitors are selling 4.0 GB disk drives, you can offer 5.2-5.6 GB for roughly the same price. Or cheaper if you're a dog.
1998 is going to be different, but KM+MR works in the lab. It is fair to assume that having channel electronics optimized for KM+MR is going to boost whatever significant capacity boost they were able to achieve in the lab, even if it gets whittled down to production level terms. If the Japanese head manufacturers come through on their boast to leapfrog anistropic MR and go straight to GMR (spin valve) then thermal decay becomes more of an issue. If Ampex is a lady with a past, she has time on her side this time around. We know that even if the market continues to diss her. |
