Some Seagate notes:
1) Volume estimates
March quarter - 7.9M units (5.6M D + 2.2M E + 200K P)
June quarter - 7.1M units (5.1M D + 1.8M E + 200K P)
September quarter - 7.8M units (5.9M D + 1.6M E + 300K P)
D-desktop
E-enterprise
P-portable
2) JIT effect
Compaq's CTO (Configure-to-order) model, which involves the channel assembly by about 9 major distributors of CPQ's workstation and server products, starts in November. This means that CPQ and its channel assembly partners squeezed out their disk drive inventories during the transition and pushed back their new orders. This move affects all the drive makers because Compaq buys from the top tier vendors. Last year CPQ bought about $1 billion of disk drives from SEG. If you knock off 40% for loss of market share and price pressures, you are still left with about $600 million worth of disk drive business, or $150M/quarter. Skew that for seasonality and it's easy to see how CPQ's JIT model alone easily cost SEG about 50-100,000 disk drives. Still not enough to salvage a bad quarter, but volume is key to SEG's ability to recover.
The Conner acquisition gave SEG a total of 50 million in disk drive capacity. Knock off 10% for obsolescence and to account for the various manufacturing transitions and SEG is still left with 45 million in annual capacity. If SEG hits its target of 8.9 million drives in the December quarter, their trailing 4 quarter production would be about 32 million. 32/45 capacity = 71%. There are significant efficiencies to be gained by bumping that up to about 80-85%, either by increasing production and/or reducing capacity. Additionally, there is a glut in platters as well as a developing overcapacity in heads so it makes sense for SEG to take advantage of that situation. By not expanding its head and platter capacity, for example, to scale with the 20% annual growth rate, SEG's 80-20 mix in heads and platters drops down to a 67-33 mix in one year, 55-45 in 2, etc...
Prospectively, SEG has to continue to push out the volume and improve its product mix to make up for the dropping ASPs in the desktop and high-end segments. That has implications for everybody's business plans. Just ask WDC. Fujitsu and SEG are slugging it out for volume in the <2.0 GB segment so WDC has to scramble to higher ground where it is going to ran smack into a heavy hitting crowd:
Maxtor (market share strategy, e.g., 7.0 GB net-priced between 6.4 and 5.1,look for their DD+free 16 bit memory chips promo this Xmas)
Quantum (MR transition complete, 50+% share of 4+GB segment)
IBM (most efficient MR head producer, 100% integrated in heads, platters and ASICs)
Fujitsu (50-50 integrated in heads,platters, and ASICs)
Seagate (80-20 integrated in heads and platters, ASICs and motors are outsourced).
The thing that makes the desktop disk drive a commodity is that the various offerings are easily interchangeable especially when you have PC makers adjusting to ASPs that are trending down fast. Sure you have a layer of high-end desktop demand where arcane stuff like access rates, transfer rates and areal density matter, but that's a tiny part of the market.
In the high-end, there's going to be a rush to 10,000 rpm and up disk drives, especially now that Fujitsu beat everybody else to play second fiddle to the Cheetahs. SEG is going to push this hard because they are the only ones with the advanced designs and the capability to mass produce the dynamic motors with the fluid bearings required to increase spindle speeds. There's an EE article that describes how SEG expects all its drives will be using these motors in 3 years probably using a 30-70% internal-external model with partners such as Sankyo, Seiko and a British outfit mass producing the motors.
In the technical scheme of things, since...
Access/Data transfer rates = Rotational speed x Areal density
SEG has to redouble its efforts to catch up in areal density because IBM is so way ahead of it with its 2.65 Gbits/in2 MR heads and its 3.0 Gbits/in2 MR heads. Though still being produced for IBM's low-volume (read: poor yields up to this point) notebooks, it's only a matter of time before IBM improves its yields and develops the read channel electronics to allow the use of these advanced MR heads at higher rotational speeds. This is the point which I believe will be fatal to SEG because it promises to trump SEG's formidable leads in spindle speed, motor technology and interfaces, assuming of course that SEG does absolutely nothing in the areal density front. Not a reasonable assumption to make since SEG leads IBM and even the Japanese, I think, in the design of MEMs to help the GMR heads move precisely from track to track.
Management-wise, there has been a flurry of moves to beef up the ranks. Aside from elevating Luczo to President, Shugart also raided 2 execs from WDC and promoted a whole host of others from within. I wouldn't be surprised if there is some serious headhunting going on targeted at the marketing department over at Quantum's headquarters, primarily to beef up the tape drive operations.
Patience is required here, along with good sources of channel reports to guide one's entry points. |
