Showing posts with label HDD. Show all posts
Showing posts with label HDD. Show all posts

2016-03-17

Disks for Data Centers

Google released a whitepaper on their wishlist for HDDs.  It's somewhat amusing that they appear to think that ECC is still used.
"An obvious question is why are we talking about spinning disks at all, rather than SSDs, which have higher IOPS and are the 'future' of storage. The root reason is that the cost per GB remains too high, and more importantly that the growth rates in capacity/$ between disks and SSDs are relatively close (at least for SSDs that have sufficient numbers of program-erase cycles to use in data centers), so that cost will not change enough in the coming decade."

2012-03-09

HDDs Gone By

old gold: Some information on HDDs gone by from the Land Down Under.

Wikipedia has a list of defunct HDD manufacturers, as well.

2012-03-05

HDD Industry Consolidation

  • Rumors from The Register circa 2012.02.08 regarding Toshiba as the buyer of Desktop drive assets from WD/HGST.
  • Verification of the rumors via WD's press release dated 2012.02.28, with additional information that WD will have to acquire Toshiba's Thailand plant in exchange.
  • Article on China approving the WD acquisition with conditions on 2012.03.02, similar to the EU's objections.  (A translation of the article was made internal to HGST; the linked article has a rougher translation.)
  • The FTC issued a proposed order to accept the WD acquisition with divestiture of 3.5" assets on 2012.03.05, apparently in response to China's approval.  However, it appears that there's an open period for comments for 30 days, implying the approval hasn't been given yet.

2011-03-12

Articles about WD's Acquisition of HGST

I was still at work in Japan when the note came out. That definitely was a shocker.  Of course it hasn't been finalized yet, but there isn't much that would stop it, eh.

Merger to Create PC Drive Giant [WSJ]
日立がHDD事業を米Western Digitalへ売却、43億ドル相当の現金と株式で [Cloud Watch]

Viviti? With the IPO we were going to be Viviti? Come on guys, surely you could've come up with a better name than that!

2011-01-14

Sued

A few days ago a coworker asked me and MRG to produce a list of products since Jan. 1, 2003 with the corresponding data channel used.  He said it was for a patent lawsuit, but didn't have additional information.  Yesterday he told me that the plaintiff is a company called Lake Cherokee, so we googled it and found the lawsuit: Lake Cherokee Hard Drive Technologies v. Hitachi Global Storage Technologies, LSI Logic, STMicroelectronics, Marvell Semiconductor, et al.  ("Hard Drive Technologies"?  What the fuck do they do besides sue real HDD-related companies?)

Turns out this has to do with something I heard about a couple months ago: the senior integration guy was looking for specs dating back to the late '90s.  This was before I joined, but I knew where to find the info.  The first item in the lawsuit shouldn't apply to HGST, because the internal channel's error measurement circuit predates '98—we have prior art.  The second one might be applicable; I don't actually know the details.

2009-06-17

Colossal MR Effect

Will the "Colossal Magnetoresistance Effect" be the next big thing for HDDs? I can't say that "CMR" is the greatest acronym ever.

2008-10-02

Fujitsu to sell to WD?

CNET had an article relaying the Nikkei report from last night.

Here's my commentary:
Fujitsu only makes mobile (2.5") and server HDDs. The 2.5" space has been getting crowded lately: HGST, Toshiba, Fujitsu were the main players, but now Seagate, WD, and Samsung have joined in and snapped up market share. It's inevitable that one would have to bow out.

WD doesn't really make server-class drives -- although they claim to, they're still 7200 RPM SATA, which is decidedly NOT a real enterprise model -- so acquiring Fujitsu would bolster their efforts in enterprise and eliminate a competitor in mobile. Sounds like a good strategy to me as long as they can capitalize on the server business.

However, merging corporate cultures is painful, as survivors of the HGST/IBM and Seagate/Maxtor acquisitions can attest to.

Incidentally, Fujitsu is denying the Nikkei rumor...for now:
http://pr.fujitsu.com/jp/news/2008/10/2.html

2008-06-01

The SSD Threat

Samsung didn't have me very worried with their cheap flash, because they were screwing themselves trying to build a market by dropping prices to unsustainable levels. Apple took them for a ride. Intel's effort is more of a concern, because they have the market clout to push flash. That Seagate's going to enter the SSD arena as well is an even bigger concern—perhaps the HDD industry's headed for some more pain in the near future. Remind me why I do this again?

2008-03-10

LSI to Purchase Infineon Hard Drive Semiconductor Business

News from tha wire.

Aw hell. What does this mean for us internal folk?

Those LSI bitches are a tenacious bunch, pulling away some top people, and now hijacking one of our previous foundry partners. Perhaps I should update that resume after all.

EETimes has some insightful remarks as well.

2008-01-01

To Power-Down or Not To Power-Down

A c|net blog post led me to an article on Ask Leo that compelled me to post a reply. Here it is:
To Roger: Not all drives use the CSS (contact start/stop) mechanism you described. All Hitachi hard drives use a load/unload ramp, which causes no friction wear to the head. (Ramps can have other issues, however.) Seagate's 3.5" drives still use CSS, last I heard, so people who use that vendor may wish to avoid excessive power cycling.

Also, most spindle bearings are FDB (fluid dynamic bearings) so bearing wear isn't nearly as much of an issue as it used to be. Modern flex tapes generally don't have fatigue issues, either, at least from what I hear being in development. (I work in the industry.)

Other remarks: What does concern me is electronics wear with power cycling, since electromigration effects are greater as transistor size shrinks (current generation SOCs are 90nm). However, risk of head-disk interaction as a result of internal contamination, application of external force, or other issues (firmware problems, failure to park on an emergency power-off, etc.) counterbalances the argument to leaving a drive on all the time, since usage of the drive raises risk of HDI. Excessive heat due to the drive running can exacerbate thermal decay of written data, as well.

All drives will experience degradation of the magnetics because of superparamagnetism, whether they are on or off. This is a practical limit on the lifespan of a drive due to degradation of the once-written servo pattern. Copying data to an HDD and leaving it in the closet for 5 years is NOT safe -- I'd prefer optical media (just make sure to use a high-quality brand like Taiyo Yuden, not the cheap stuff).

In the end, because of the complexity of HDDs, the differences in models, and the difficulty of extracting probabilities of failure, whether to power down or not is still a matter of preference. I shut off most all of my HDDs when not in use, since I view the mechanical and magnetic risks to be greater than the electronics ones. Thus far, I've had good luck. If one chooses to leave a drive on all the time, definitely make sure it's well cooled. Multiple backups over multiple drives is the safe thing to do. Also, don't trust what the manufacturer says about drive life -- the one oddball failure on the Weibull projection could be the one with your data on it.

2006-10-22

HDD Speaker

If one's into low-fi, apparently someone cooked up a way to mod an HDD voice coil into a speaker. Demented, but it comes with a video....

2006-10-20

News

Invention firms

Google earnings

CNN Money 25 rules to grow rich by

Internet addicts as ill as alcoholics? - Well, it's not a chemical addiction, anyway.

Wisdom of non-crowds - Yeah, crowds are dumb. Are non-crowds any better?

Samsung's hybrid HDD

Single pixel digicam - Interesting, but of questionable practicality.

2006-03-02

Hard Disk Info

USByte has an interesting explanation of HDD technology. It's extremely simplistic, but is therefore good information for someone who knows nothing about how a HDD works.

Um, the following is work-related, but is kinda interesting. Regarding the so-called "Wallace spacing loss":
Clearance Measurements - References the patent by Klaassen, et al, #5130866.
Head reliability of AMR sensors based on thermal stress tests - A bit more detail regarding the spacing measurement.

2005-12-21

Seagate to buy Maxtor

Coverage from:
New York Times via c|net
Reuters via Yahoo!
Mercury News

With "Quaxtor" gone, the HDD industry is left with Seagate, HGST, WD, and Samsung in Desktop 3.5" and Seagate, HGST, Toshiba, Fujitsu, WD, and Samsung in Mobile 2.5".

2005-12-14

Serial ATA and Microsoft's OSes

Because it's relevant to me at the moment:
Serial ATA in the Microsoft Operating System Environment

Whew -- the HDD I thought I'd killed actually only had a mangled pin (caused by a connector with a blocked pin for the key pin, but no tab for the recepticle key). A bit of soldering and it's as good as new.