Fujitsu tells WD the deal's off Fujitsu did not give Western Digital a Christmas present - in fact it said 'No deal'. The two companies have been talking about Fujitsu selling its loss-making hard disk drive business to Western Digital. Fujitsu's president Kuniaki Nozoe said there was zero chance of a deal because Fujitsu wished to sell its Japan-based and …
I can't imagine what the outcome would have been to throw the people who gave us the MPG/MPF series fiasco into a quality manufacturer such as WD. I wonder would the result have been the same as the Seagates being manufactured in China by the Maxtor people who gave us the stunningly reliable DiamondMax 9?
I have a long memory and have avoided Fujitsu drives since the late '90s. Remember this one? Drive in the freezer to allow the Cirrus Logic controller that had been baked beyond recognition (damages the dielectric film between the encapsulation and the die, releasing the phosphorous flame retardant in the encapsulation which rapidly becomes phosphoric acid and eats away the metal interconnects allowing the ions to migrate and form shorts between pins) to contract back to some semblance of normality, hopefully removing the resultant internal shorts for long enough to get your data off. Maybe. It became apparent that a long ATA cable and a freezer with a fat seal was required.
PCBA swaps were infeasible as there were so many damned variants of the things, they all had different (hidden) sector reallocation tables and the chances were they had ALL been reflowed with the wrong profile. Worse, most of the bloody Dells, Compaqs, HPs and even some IBMs produced at that time had the buggers in 'em and they, to almost a unit, failed at about the same age. Idiots. Obviously the process engineering department's week off. That, or Cirrus didn't translate the data sheet for them, so quite naturally Cirrus got the blame for their "poor choice" of encapsulation. Note that Seagate, using the same controller (on the ST3x "rubber johnny" variants, I believe), didn't see anywhere NEAR the same failure rate. Draw whatever conclusion you wish from that.
At this rate, only the Samsung Spinpoints are going to be seeing the inside of my machines. Hitachi Deathstars and Travelstains have too many memories attached to them. Roll on cheap, reliable SSDs. Even the Chinese couldn't botch up one of these... could they?
I hope that these storage media companies begin manufacturing 2TB disks before they fail financially and fire the R&D department.
The obsessive-retentive part of me thinks that 2TB is "perfect" because that is the addressing limit for computers with parallel ATA/IDE interfaces.
The 2TB hard disk could be the next 1.44MB floppy disk or the next 2GB SD card, especially for embedded systems. It would be a mature part for a stable old platform that just can't go any bigger, but doesn't really need to go bigger.
I agree with you to a point. In one of my old retail jobs (1997), we sold IDE hard drives from WD, Fujitsu, and Samsung. Never had a single WD fail, but many of the Fujitsu and Samsung drives failed on their first use (bad sectors during format, with more bad sectors during each successive format). At my next job (1997 to 2004), we used Seagate SCSI drives exclusively in our servers, despite the fact that the drives consistently had a 33-50% DOA rate (either wouldn't power up or wouldn't be detected by the controller).
My first personal computer (1992) had a Seagate IDE hard drive that failed after just over a year. Since then, I have personally had IDE drives from Maxtor and WD fail, both DOA and after use. Last year, I built two servers using WD's 150GB Raptor drives, and 2 out of 8 were DOA (they would spin up, but wouldn't be detected by the RAID controller). I've used and seen enough drives from enough companies to know that the quality of all drives is next-to-nothing nowadays, and that you simply cannot equate a manufacturer's name (or reputation) with quality. Back in the 90s you could, but not today. Having said, I'll never use or recommend a Seagate, Samsung, or Fujitsu drive. You know what the they say about first impressions.
@Chris C: Funny thing about the new Raptors and, to a certain extent, the RE2s. Some of them simply will not negotiate with SATA 150 controllers without the jumper in place (pins 5&6 of the feature block off the top of my head), giving the impression of a DOA drive, yet some will work perfectly without it. I have an RE2 sitting on the bench right now that refuses to work correctly (LBA timeouts, looks terribly like bad sectors) with a Via (ick!) based southbridge, yet sticking it into my test box and running WD's own diagnostics, along with smartctl and testdisk on BSD, shows the thing to be perfectly healthy. One wonders just how many of these "failures" are due to poor HBAs, cabling and power issues. When designing servers myself, I usually use HBAs such as the Highpoint or Areca to ensure this doesn't present a problem.
That said, I'm sure you've covered all of this yourself, so I don't doubt for one second that you have had actual failures. I found the WD Proteges (usually fitted to OEM kit. They're not that common) to be probably the worst series for returns, although it was nowhere near the level of incompetence you suspect when a whole batch of Seagates (this is recent) with the place of manufacture indicated as China ALL coming back as faulty within days. As you say, there are failures with all manufacturers these days, as Storage Review's database clearly shows, but I think it's fair to eschew a particular brand when they show the levels of incompetence that Maxtor-Seagate and Fujitsu have in the past, especially when it makes you look incompetent yourself to the non-technical.
Again this is personal experience, but I have seen a total of one failed Samsung Spinpoint, probably due to not having used or shifted too many of them. This was due to someone thinking the breather was a "reset button" and prodding it with a pin. No matter how idiot-proof you try to make something, nature still seems to be ahead with the idiots :o)
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
