Cool, Stitch. Great info.
Re: Japan, Inc.
I didn't know exactly how advanced Fujitsu's disk drive R&D was, altho, their aggressive work with Tunneling MR (10 Gbits/in2 +) did suggest that they also have an equally aggressive GMR/CMR program.
Anyway, my understanding is that TDK/SAE, Yamaha and Fujitsu all licensed their basic MR designs from IBM (read: use yen to buy dollars to pay royalties). By bringing Toshiba's GMR design to market more quickly, this gives them an opportunity to reduce their royalty payments to IBM. It also allows the Japanese to avoid paying additional royalties to Headway, the vendor with the most practicable dual element MR head design (already licensed by IBM, WDC and SEG; field-proven as the head design of choice used in SEG's 2nd generation Cheetahs). IBM has demonstrated that this design can extend conventional AMR to 5.0 Gbits/in2.
October 16, 1997 - IEEE Seminar
"5 Gigabits Per Square Inch Recording Demonstration with Conventional AMR Magnetoresistive Heads and Thin Film Disks" www-physics.ucsd.edu
The big question mark is IBM. Right now, Big Blue can...
1) Move quickly to GMR. Doing so means that the Japanese will move more quickly to GMR also. There goes that conventional MR royalty stream.
2) Move more slowly to GMR, perfecting the processes, while extending conventional MR to 5.0 Gbits/in2. I think SEG's priority right now is to make desktop disk drives faster, i.e., rotational speed x linear density = data transfer rate, which is why they're bringing 7200 rpm drives to the desktop. IBM also has 7200 rpm desktop drives coming down the pike, altho, they're initially going to be using it in 10+ GB drives while SEG's 7200 rpm drives will be at 4.6, 6.9 and 9.1 GB -- 2 to 4 platter drives using 2.3 GB/platter MR heads.
If IBM and SEG focus on speed, it would seem to be foolhardy for the Japanese to try to move too quickly to GMR because they risk poor market acceptance, the same type that plagued QNTM's premature move to MR.
Here's one early review in which the Medalist Pro basically smokes the IBM Deskstar, QNTM's Fireball SE and Fujitsu.
zdnet.com
SEG Medalist Pro
"Simply the biggest and fastest drive out there - bar none."
PROs: 9.1 GB capacity; fastest EIDE drive on the planet; 5.4 cents/MB
CONs: May be hard to find individual drives until March of next year.
QNTM Fireball SE
"Big, fast drive but when it comes to price and performance, it can't beat the Seagate."
PROs: Fast performance; 8.4 GB capacity; 6.8 cents per megabyte
CONs: Can't touch the speed, capacity or even price of the Seagate
Fujitsu
"Not a remarkable drive, but its low price makes it appealing."
PROs: Affordable and available now.
CONs: Slow overall throughput; 7.8 cents per megabyte.
At this point, the bottomline is I don't think IBM has the luxury of moving to GMR all that quickly, because, if this early review is any indication, SEG's focus on speed matched by competitive capacity and aggressive pricing will be the elements of a winning desktop strategy for the next few quarters. That, IMO, forces IBM to follow SEG's cue because I just think that SEG's price and performance proposition is too compelling. It's just too bad that the industry supply and demand imbalance basically forces SEG to give up the natural premiums of a performance-leading disk drive because it has no choice but to go after volume, volume and then more volume.
March quarter - 5.4M desktop + 2.2M high end + 300K = 7.9M units
June quarter - 5.1M desktop + 1.8M high end + 200K = 7.1M units
Sept. quarter - 5.8M desktop + 1.6M high end + 400K = 7.8M units
Dec. quarter(est) - 6.9M desktop + 1.6M high end + 400K = 8.9M units
Note: Assume ASPs and Gross Margins in the gutter.
Lehman's estimates of channel inventory with about 2 weeks to go:
Desktop - ~13 weeks
High end - ~10 weeks
Note: Some analysts have been noting the build up of inventory among the PC makers even as they transition to BTO/CTO. That aggravates an already funky situation for the drive makers and their suppliers. Watching all that inventory overhang rip through the supply line is going to be like watching several accidents happen in slow motion at the same time.
My understanding also is that SEG has been prepping the OEMs for this technology transition for at least 1 1/2 years now, particularly the fact that they are using fluid-bearing motors, instead of ball-bearing motors, for the very first time. You probably know that SEG has been single-handedly building the supply line for these proprietary motors (part internal, part outside suppliers like Sanyo, Seiko, etc) as part of a 3-4 year effort to transition all its drives to this platform.
Fluid motors put a faster spin on the desktop drive
techweb.com
Would you know how IBM and Fujitsu stack up competitively in this area?
Thanks, Stitch.
Gus |
