Seagate forced to think about the next small big thing

Comment Western Digital has pipped Seagate in the small form factor (SFF) areal density hard disk drive stakes with today's 1TB Scorpio Blue, which, in My Passport Essential SE form, compares to Seagate's 640GB FreeAgent Go. However the Seagate product is a two-platter one, meaning 320GB/platter, whereas WD's has three.

That means that WD's product has a unit depth of 12.5mm, rendering it unsuitable for standard notebook computer SFF drive bays, which require a 9.5mm drive depth measurement. WD has 9.5mm deep Scorpio Blue drives; these though, have two platters and top out at 500GB, meaning 250GB/platter.

The 1TB and 750GB WD drives spin at 5200rpm, whereas the 500GB Scorpio Blue model spins slightly faster at 5400rpm.

WD's faster Scorpio Black notebook drives spin at 7200rpm and hold up to 320GB. Seagate has a Momentus 7200.4 7200rpm drive offering 500GB. It's a two-platter, 250GB/platter model fitting the 9.5mm depth requirement. It also has the Momentus 5400.6 drive spinning at 5400rpm and offering up to 500GB.

The 320GB/platter drive is, we suspect, or will be a Momentus 5400.7. Stifel Nicolais analyst Aaron Rakers says: "Seagate had announced that it has delivered its 320GB/platter, two-platter 2.5-inch solution to its US retail partners and a ramp is anticipated in the September quarter."

So Seagate has the highest-density internal notebook drive spinning at 7200rpm while WD has the highest-capacity external SFF drive, with the highest areal density, at 1TB. It also has its 10K rpm 2.5-inch, 300GB, SATA Velociraptor enterprise drive

Where do we go from here?

It seems intuitively obvious that WD will take the 333GB/platter technology from its 1TB and 750GB Scorpio Blue drives and scale down the capacity, until it can be reliably read and written to at 7200rpm, hopefully producing a 500GB or greater Scorpio Black to match or surpass the Seagate 7200.4 Momentus. Then it can do the same with the Velociraptor and possibly produce a 400GB product spinning at 10K.

WD can also take the 333GB/platter technology and produce sub-750GB 2-platter Scorpio Blue products with it, at possible 666GB, 333GB and 166GB levels, the latter two with just one platter and headset, meaning cost savings.

Because of its areal density leadership here, WD can turn the screws on Seagate.

As a response, Seagate may produce a 3-platter FreeAgent Go with its 320GB/platter technology and so have a 960GB product: it wouldn't have the same shelf appeal as a 1TB SFF drive though. Seagate would need four platters to break the 1TB barrier, with a 1.28TB product. Another platter and another head means more cost, which would be a disincentive to doing this. It's not obvious that Seagate will go this route.

So the second intuitively obvious development would be for Seagate to move to the next level of SFF perpendicular recording technology and attain an areal density of... what exactly? WD bumped up its 250GB/platter Scorpio Blue areal density by a third to get to 333GB/platter, If Seagate applied a 33 per cent areal density uplift to its 320GB/platter technology it would arrive at 425GB/platter. This would give it an 850GB/2-platter drive spinning at 5400rpm.

Were WD to add a 33 per cent uplift to today's 333GB/platter Scorpio Blue drive it would get to almost 430GB/platter. Seagate's strategists must surely be thinking that that they have to at least match WD areal density levels, meaning 430GB/platter or, and this could be so enticing, go for gold and 500GB/platter.

That would produce a 1TB, 5400rpm, 2.5-inch drive. Ramp that down to acceptable read and write rates at 7,200rpm and it would have WD beat at both 5400rpm and 7200rpm capacity levels, at the internal notebook drive 2-platter product level, and in the retail shelf SFF drive capacity stakes.

It means a 56.25 per cent jump from the current 320GB/platter technology Seagate has now. Can that be done? Such a jump will take a lot of development and significant time. It would be the only certain way for Seagate to retain its lead over WD, as SAS drives and solid state drives erode Fibre Channel drive's dominance in the enterprise space, and 2.5-inch spindles are used to get over the access density limits of 3.5-inch drives.

So far WD has not adopted SAS. Well... suppose it were to do just that. No company spokesperson has said it will and previous WD comments about SAS merely note it as a growing factor in enterprise hard disk drive circles. Reading the runes here could lead one to suppose that WD could take its SATA SFF areal density superiority, add SAS interfaces to the mix and really give Seagate a hard time - Savvio here we come. That prospect, if real, surely means Seagate will find a 500GB/platter 2.5-inch drive areal density a compelling target.

What a juicy prospect. ®