WD to acquire Hitachi’s hard-drive business

Consolidation is afoot in the hard-drive business! This morning, Western Digital announced its imminent purchase of Hitachi Global Storage Technologies, Hitachi’s mechanical storage subsidiary. The deal, which is expected to close in the third quarter, will see WD cough up $3.5 billion in cash and 25 million of its shares worth a combined $750 million—that’s just under $4.3 billion altogether—for its former rival.

WD and Hitachi still need to get the nod from regulatory entities, but their boards of directors have already approved the deal. If the transaction completes without hitches, Hitachi will end up owning a cool 10% of WD shares, and Hitachi GST CEO Steve Milligan will become President of WD. (John Coyne will, however, remain WD’s CEO.)

A story by the Wall Street Journal sheds some additional light on the deal. Reportedly, Hitachi is moving “away from consumer electronics toward large industrial projects,” and the company will invest much of the cash from the sale in “infrastructure businesses such as railway and electric power.”

As the WSJ points out, Hitachi GST consists of the remnants of IBM’s hard-drive division, which was sold to Hitachi in 2003 in the aftermath of the 75GXP failure fiasco.

Comments closed
    • HisDivineShadow
    • 9 years ago

    Wow. Didn’t see that one coming. Though WD did try to buy Seagate a while back. Or rather, I think Seagate tried to sell to WD, I think. Hitachi selling out to WD is not as big, but it is an interesting highlighting of the fact that hard drives are yesterday’s news and the bit players are pulling out in a big way. Hitachi was never as huge in this as WD or Seagate, so they must have decided the hassle just wasn’t worth the trouble. By and large, it appears the only thing WD gets from this is a slightly bigger piece of the pie. The consumer gets higher prices on hard drives with reduced competition since Hitachi was more or less a value brand of hard drive by this point.

    I just don’t think the Deskstar name ever really got all of its luster back following the Deathstar Fiasco.

    The writing is on the wall. Prices are going to continue to drop on hard drives and SSD’s and the lower the latter go, the lower the former will be forced to go. Especially as size to dollar ratios improve on SSD’s. Meanwhile, failure rate is going to improve on SSD’s and stay about the same on hard drives, eventually leading to a perception that SSD’s are flatout the better buy due to reliability.

    Better to get out while the getting out is good, I guess. If they wait until SSD’s own the market, they won’t get as much cash. In other words…

    WINNING!

    • omf
    • 9 years ago

    Well, let’s see…

    Conner got bought by Seagate in 1996
    Quantum sold off to Maxtor in 2001
    IBM sold off to Hitachi in 2003
    Maxtor got bought by Seagate in 2005

    Those are all the players I can remember…

    • continuum
    • 9 years ago

    I just hope WD keeps Hitachi’s enterprise firmware guys, ’cause yeah, I like my harddisks that WORK with RAID instead of randomly dropping out like flies…

    • albundy
    • 9 years ago

    so i guess its sammy vs WD vs seagate now. i originally predicted that seagate would own them all before the fall of mechanical drives, but it could be WD. its really exiting to see who will be the biggest loser and own the monopoly of these old clunkers.

    Does this also mean that WD will change my warranty on the 2TB hitachi drive i just bought? i wouldnt mind wd’s 5 years instead of the 3 years that i got.

      • AlvinTheNerd
      • 9 years ago

      The triumph of SSD over mechanical is far from assured.

      For one, there is no guarantee that node sizes will continue to get smaller. There is Moore’s first and Moore’s second law. The second being that the cost of new fabs will increase exponentially. Eventually this catches up and no company in the world can afford to make a fab for the next node. Further, node decreases use to depend on the application end. It was just a process of trying to get smaller amounts of dopants on to silicon with percision. But now, new materials have to be invented for each node decrease. Eventually this also catchs up on you.

      No one can say how far Moore’s Law can go, but everyone, even Moore himself, agrees it can’t continue forever. When it slows, SSD advancement might be stopped before overcoming mechanical in price per GB.

      Second, flash technology is getting more and more fragile with each node decrease. The number of rewrites decreases with every generation. While this still isn’t an issue, eventually this might keep SSD’s as a niche fast drive and not something that reliably stores the mass of information that HHD’s do now.

      Its two technologies, and SSD is advancing more quickly at this point, but there is no way of saying that this will continue nor that HHD will not find their own ramp up in performance and/or capacity.

        • UberGerbil
        • 9 years ago

        I agree with all your caveats, and certainly at the enterprise / datacenter / cloud end of the scale, where enormous capacity is the essence of the business, it’ll be a long time before HDs die out. That definitely will be the last stand of the HD dinosaurs, and it may not be flash-based SSDs that are the meteor that ultimately kills them: something else “solid state” may supplant spinning disks and NAND both.

        But it’s also pretty apparent that SSDs are going to replace hard drives as the everyday storage for ordinary users. SSDs already have achieved the minimal “walking around” capacity for most people: if SSDs didn’t cost so much, almost every laptop would already be sold with one — and most desktops, too. They’re inadequate for home media servers, of course, but that’s still a relatively small niche and for the mainstream one that may ultimately be filled by their cable company’s VOD service rather than by a home NAS. The future looks more like relatively small amounts of data held locally on SSDs and much larger amounts pushed out the the “cloud” where HDs remain, with various content / application providers (Netflix, Steam, Amazon, etc) mediating between them.

          • cynan
          • 9 years ago

          Right now, and for the foreseeable future, SSDs are only relevant to certain high-IOPS applications and the computer enthusiast. Someone looking at a $500 workstation, whether to populate business cubicles or the desks of home users, will not care about shaving seconds off of booting Windows or starting up MS Office, and certainly not at the high increase in $$.

          For enthusiasts, it does not make sense to limit yourself to either an SSD or mechanical drive, with the obvious best-of-both-worlds solution being to buy a smaller, more cost effective, SSD for a boot drive, while having secondary mechanical disks for most documents, media, backup, etc. The only area where you need to choose is when looking at ultra portable notebooks (13.3″ or less) as most only have room for a single drive – but even this may soon change with the introduction of bootable mini PCIe flash drives…

          In a couple of years, when SSDs become more cost effective, they will see more market saturation, no question. For example, when prices get competitive enough, motherboard manufacturers may start to bundle smaller SSD boot drives with their enthusiast lines – perhaps even soldering to the boards for cheap packaging… But regardless, none of this suggests that mechanical drives as a means of mass storage even for the end user, is going any where any time soon. Nor that SSDs will even outsell mechanical HDDs in unit quantities for many years to come.

    • demani
    • 9 years ago

    Plus Hitachi has a set of partners they sell through as well (G-Techtonology is off the top of my head, but I know they have others. While individually they tend to be pretty small, as a whole they add up to a lot of extra distribution channels. Besides, for Hitachi- while they seem to have been of fine quality, they weren’t ever going to take over the market. I guess the major players are just going to be Seagate (and they have their own reputation that needs to rebuild),WD, and Samsung? I’d guess Samsung might be the next to go though they make so many other things that maybe they’ll buy Seagate or WD.

    And for all the rumble about SSD it still has longevity concerns, and capacity concerns (4TB mechanical may be coming soon, but 1TB in SSD is still going to cost at least an order of magnitude more). There will always be a place for HDD.

      • UberGerbil
      • 9 years ago

      And Toshiba, who bought their HD business from Fujitsu.

      I think Samsung is really only in the HD business at this point as a way to cultivate relationships for their SSD business. Remember that a lot of other SSD companies get their NAND from Samsung, so they have a much bigger stake in the SSD future than they have in the HD past (and in that sense they don’t really care who wins there as long as the winners use Samsung flash, but the margins are higher on all-up SSDs than they are on bulk flash). OEMs, meanwhile, prefer to deal with fewer component suppliers that each can provide them with more options. Dell isn’t going to buy [i<]just[/i<] SSDs, at least not yet, but -- all else being equal (which of course it never is) -- they'd rather deal with one supplier who can offer them both HDs and SSDs than separate suppliers for each. So Samsung keeps a toe in that water, at least for now. At some point Samsung probably sell off the mechanical part of their drive business to one of the other remaining players; I wouldn't put it past Samsung buying a drive company, but it would be to get their retail SSD marketshare, not the spinning disk part. And actually, if I was Samsung, the SSD-related company I'd be trying to buy is Sandforce. Though I'm sure they have to put their name down below a lot of well-known others to even get that dance started.

    • gamoniac
    • 9 years ago

    [quote<]The market is clearly moving to SSD's, ...[/quote<] I just bought a 1.5TB for USD$70. That is 4.7 cents per GB. Compared to roughtly $1.9/GB for SSD, that's 40 times cheaper. Except for the rich few, I don't think many households have 1.5TB worth of SSD for storage. It will be quite a few years before SSD price/GB catches up to HDD. At the meantime, from the business perspective, every dollar you don't make if another dollar for your competition to grab, and use it to invest in technology of the future (like SSD). WD's objective might be short term, but no body can get too far without solid short term goals. Of couse, as in all investment, every move is risky. $4.3 billion is not a small sum, so I can see the other side of the argument as well. Cheers. P/S: and if you chain a couple of 1.5TB (or even 500GB HDD) in RAID, the performance is nothing short of remarkable. Especially for use in VM and SQL environment. I am getting 246MB/s and 244MB/s for sequential Read and Write with 2x500GB Seagate Barracuda RAID-0. Of course, back up is a must.

      • Chrispy_
      • 9 years ago

      There’s not a lot of profit in the mechanical disk market. Competition is fierce and the end product is relatively cheap with a small margin.

      When a 1.5TB drive lasts most people for the entire lifespan of their computer, if not longer, getting back $4.3bn is hardly a short-term investment.

        • magila
        • 9 years ago

        That depends one just which part of the market you are talking about. The enterprise market has quite good margins and demand for enterprise class hard drives continues to be quite strong.

      • indeego
      • 9 years ago

      Hopefully you mean non-production VM/SQL with RAID0. (I can’t really see a situation where a nonproduction DB/SQL would benefit from RAID0, period.)

      I’d rather go with an enterprise SSD for both of those, (Good luck with 60K+ IOPS with RAID0 mechanicals) and would never consider RAID0, ever, ever, ever.

      The only situation I would use RAID0 would be in a scratch disk situation, that only held temporary data that was always considered noncritical. If you want fast performance just look into a faster controller/ I/O with SSDs.

      The downtime/hassle of the increased chance of failure is simply not worth the slight performance gain. If speed is important to you, then pay for it.

        • gamoniac
        • 9 years ago

        I know I should have been more specific, my RAID-0 is not for production environment. For enterprise-level data, I can’t imagine how SSD can provide the capacity needed in many cases.

        In my case, I use it for personal develpoment use. My RAID-0 is backed up nightly. All things considered, with two kids, I simply can’t justify spending for even 1TB of SSD storage. I am fully aware of the risk of RAID-0, but I wouldn’t call the gain marginal, at least from what I can see and feel. I will keep my fingers crossed, and might still switch to RAID-1 if I get burnt. Thanks for the advice.

      • sjl
      • 9 years ago

      Sequential I/O is easy. Trivial, in fact. SATA drives do it very well indeed; if all you need is high sequential throughput, then yeah, stripe together a bunch of SATA drives (even use RAID 5, if redundancy is necessary), and they’ll fly.

      That’s not where the interest of corporate environments lie, though. Random I/O is the big problem there. Consider databases – most of the time, you’ll have the disk heads flying all over the place to retrieve data, and it turns out to be the seek time that’s the limiting factor. That’s where SSDs really shine.

      I work in backup and recovery. The backup systems I manage (Tivoli Storage Manager) have a database at their core. If I were given free reign, and a blank cheque, to build a new backup system, I’d throw the TSM database, and its logs, onto a few SSDs, and the disk storage pool onto a honking great big SATA RAID set. Pre-initialise a large bunch of files, so they’re sequential on disk, and you have a backup system that’ll pull in data as fast as it comes through the network (often not all that fast), lay it out neatly on disk, and then shovel it out to tape at streaming speeds (LTO5: 140 MB/s native, somewhere in the vicinity of 300 MB/s compressed – no way will most backup clients manage that rate.)

      Both rusty iron drives and SSDs will have their place for a long time to come; spinning rusty iron drives for sheer volume of storage, and SSDs for applications where super fast response to random I/O is critical. The only reason the HD market is shrinking is that we’re hitting capacities now where most ordinary people simply can’t fill the drives up. What would you do with, say, 5 TB of storage? I don’t think I have more than 1 TB of useful data; probably nowhere near it. (ignoring DVD rips and similar.)

        • gamoniac
        • 9 years ago

        I work for a huge financial company where tens of thousand of 15K RPM HDD are still spinning away everyday. To convert all operations to SSD will take years. And I agree that random IO is where SSD shines, and I also wish a blank check to upgrade our system. The truth is, choosing between system upgrade and keeping employees, I am glad that my company chooses the latter.

        DB backup/restore and file copying would be perfect scenarios for sequential IO, I would think. But you’d probably know better in that department.

        For personal use, I have 300GB of data (projects + media). A couple of 1.5TB HDD serves me well for storage (one live backup, one for off-site storage) — low cost, and not having to worry about cleaning up for space frequently.

        Don’t get me wrong, SSDs are awesome. I would get a couple if I could. I am just unwilling to pay 40x more per GB at the moment.

          • sjl
          • 9 years ago

          Database backup is, indeed, a perfect scenario for sequential I/O. But that’s not what you spec a DB to do – it’s production use, lots of queries, that you spec it for, and that’s where databases start seeing lots of random I/O, and why you might spend the money on SSD for the most critical DBs. You wouldn’t convert everything – too damn expensive – but if there’s one database that’s a bottleneck, and it’s not too big, the price might well be worth it. Cost/benefit, the driver of all enterprise decisions in IT (which is as it should be.)

          File copy – depends. Lots of small files is a pathological case; nothing will handle that well, because of the OS overhead. It’s also a question of where the files are stashed on the media, and what order they come up. Relatively few large files, laid out sequentially on disk, is my dream scenario; pity it doesn’t come up often.

          I do hear you on the price issue; that’s the nub of the problem. If SSDs were no more expensive (or only, say, double the price) on a byte-for-byte basis, hard drives would be dead. Nobody would buy them. If prices come down, there’ll be a squeeze on the hard drive manufacturers, but at the moment, the manufacturers of SSDs seem to be more interested in increasing performance and letting the “slower”, presumably cheaper, chips fall by the wayside (might just be a simple case of performance being easy to improve, but bulk capacity not – I don’t know the economics of SSD manufacturing.)

          I’ve splurged on an SSD; it’s going to be my OS drive, with my main data (photographs, and what have you) on rusty iron. That should give me a very nice performance boost, without spending too much. And that’s what the SSD market is all about at the moment – finding the major performance bottleneck in storage, and removing it (or making it a little bit less of a bottleneck), without massive, costly full-scale overhauls. Kinda like memory – you have your slowest RAM, which is relatively cheap, then L3, L2, and the most costly L1 cache. Not much of the latter, but it’s there in the most critical area for when it’s needed.

      • Vasilyfav
      • 9 years ago

      Congratulations on not understanding the purpose of SSDs.

        • gamoniac
        • 9 years ago

        I think you are missing the point of the discussion here, which is that, as opposed to what some think, HDD will be around for a little while until SSD catches on.

      • Anomymous Gerbil
      • 9 years ago

      “blah blah that’s 40 times cheaper blah blah”

      That’s a tired old argument.

      When you can buy a boot SSD for about $100, who cares if the equivalent size HDD (if you could evben buy one that small) would cost a lotless? For the vast majortiy of people* it’s just not enough money to worry about, given the performance benefits an SSD brings.

      * [citation needed]

    • PrincipalSkinner
    • 9 years ago

    Why not invest in SSD R&D instead?

      • LoneWolf15
      • 9 years ago

      You mean they can’t do both?

    • sschaem
    • 9 years ago

    Hitachi paid 2.05 billion in 2003 to IBM, sold for 4.3 billion 8 years later.

    I’m not sure how this makes any sense, maybe WD want to sell it back to IBM in 8 years for 9 billions ?

    On the surface this deal makes no sense.

    WD stock has been plummeting over dropping demand for mechanical HD, how having more capacity help ?

    Unless I’m mistaking Hitachi doesn;t expand WD product line in any shape or form, just more of the same. production capacity, something it seem WD doesn’t need.

    Or does Hitachi have some ‘diagonal’ technology that will double density? or some other tricks to reduce reliability per bits?
    And so, WD think this its worth 4 billion ?

    So I can see how hitachi would want to sell at this price… but whats in it for WD? anyone ?

      • sircharles32
      • 9 years ago

      “but what’s in it for WD?”

      I’d guess the enterprise tech: SCSI and SAS drives at 10K and 15K RPM.
      Those are much higher margin devices.
      Also, couldn’t hurt to maybe take some of that tech and spruce up the Velociraptor line.

      It looks like WD has gone full circle, since they dumped their SCSI tech many years back.

      Hey, it’s all speculation at this point, right? Until we see actual products, based on a mingling of technologies, we’ve just got our best guesses.

        • Chrispy_
        • 9 years ago

        SCSI feels like it’s dead these days. Clever SAN controllers can saturate 4GB/s fibre connections using just SATA disks. More spindles is a better way to get increased bandwidth and IOPS. It’s cheaper, more efficient, cooler, quieter, etc.

        For anything that needs more IOPS performance, SSD is the obvious choice – and typically high IOPS requirements aren’t *usually* large capacity requirements. SCSI is now the middle-option, but it offers only a little more performance for significantly more cost. Don’t forget that SCSI capacities are much lower than SATA too, and not dissimilar to SSD’s these days. I don’t see any 3TB SCSI drives and I’m not paying over a grand for a 600GB SCSI disk…..

      • grantmeaname
      • 9 years ago

      You realize the value of any company changes over time, right? And that it’s not always at a constant rate?

      • UberGerbil
      • 9 years ago

      It’s not about acquiring more [i<]capacity[/i<]; it's about acquiring more [i<]marketshare[/i<]. It's quite possible to increase your revenues even as the overall market shrinks so long as you increase your share of that market faster than it is shrinking. HDs, as others have noted, are a low-margin business so archiving profits is all about taking advantage of scale. It's pretty obvious WD is going to be one of the Last Men Standing in the HD business, and it will remain a business for a long time thanks to enterprise data centers and trans-enterprise "clouds" even as SSDs take over for individuals, but they won't get there by just allowing their remaining competitors to consolidate against them. If Hitachi was going to fall they want it to fall into their arms rather than, say, Seagate's. And (again, as others have noted) there may be more than just market-share at work here: Hitachi is heir to some of the IBM storage patents, and those may be worth quite a bit to WD.

        • JustAnEngineer
        • 9 years ago

        Buggy whips.
        [url<]http://www.youtube.com/watch?v=p7rvupKipmY&t=08m21s[/url<]

    • Sargent Duck
    • 9 years ago

    Interesting. The market is clearly moving to SSD’s, and although wide spread adoption won’t be for another 5 years or so (just guessing), mechanical hdd’s will slowly be phasing out. Yes, they still remain the best $/GB, but SSD’s will be catching up quickly.

    I wonder why WD felt they needed to do this? I’d have spent that money investing in SSD tech…

      • Deanjo
      • 9 years ago

      [quote<]Yes, they still remain the best $/GB, but SSD's will be catching up quickly.[/quote<] I doubt we will be seeing 2-3 TB SSD drives for a couple hundred dollars anytime soon (5+ years at least).

      • LoneWolf15
      • 9 years ago

      So, tell me where I can get that 1-2TB SSD for my TiVo.

      How about my Windows Home Server? I’ll need a bunch of 2TB SSDs there too, and they need to be about $100-150 each.

      Finally, I need a few more of those 2TB ones for my VMWare server at work.

      Everyone with your question seems to forget that just because WD is investing in platter storage, it doesn’t mean they can’t invest in SSD technology at the same time.

      Also, for the near term, you’re not going to get large, reasonably priced SSDs for data storage. SSDs have their place in the market, but that doesn’t mean that platter drives do not.

        • nanoflower
        • 9 years ago

        You are correct in saying that WD can invest in SSD technology at the same time they expand their mechanical drive capacity by purchasing the Hitatchi Global Storage Technologies. The problem is that they are paying a LOT of money for something that doesn’t seem likely to change their product line that much.

        But the biggest potential drawback as I see it is all of that money isn’t available to put into SSD development. Now maybe WD has the kind of bank account that can allow them to spend heavily on developing SSD technology while purchasing another HD company but that doesn’t seem likely. I’m sure there’s some reason why WD sees this as a good business decision but I’m not quite sure what it is.

          • Anomymous Gerbil
          • 9 years ago

          “Now maybe WD has the kind of bank account that can allow them to spend heavily on developing SSD technology while purchasing another HD company but that doesn’t seem likely.”

          Without anything even resembling figures to back up that statement, it’s not really worth making.

    • dashbarron
    • 9 years ago

    I was wondering if you were going to make an article on this. I’m pretty dissapointed because despite the “Death Star” their drives have improved and I’ve used one for two years now without incident. I think their quality is definitely superior to WD.

    • NeelyCam
    • 9 years ago

    Is WD going to start making DeathStars now? Or have they already been doing that with the Green line…?

      • NeelyCam
      • 9 years ago

      .. I must have hit a nerve

        • Meadows
        • 9 years ago

        No, your comment is just stupid.

          • willmore
          • 9 years ago

          As the owner of three dead Greens, I think he was spot on.

            • Meadows
            • 9 years ago

            Serves you right for choosing a “green” product.

            • cygnus1
            • 9 years ago

            i disagree, got 8 of them running in my home server and they’ve been rock solid for over a year.

            @willmore – what was the situation where you had 3 die?

            • willmore
            • 9 years ago

            I had 4x 2T drives in a RAID5 in a server in the basement–temp stable, no vibration, etc. Two Seagate 2T Barracuda LPs and two WD 2T Greens. One green died–sent back, replaced. The other one died–sent back, replaced. The first one *died again*–sent back, replaced. At this point, I saw the writing on the wall and bought two more 2T Barracuda LPs and swapped out the WD drives. If I had known these things were so unreliable, I’d have made it a RAID6.

            FWIW, I’ve been running RAID arrays for years and this is my first set of drive failures.

            • hapyman
            • 9 years ago

            Any chance your basement is too humid? Strange if they keep dying on you like that.

            • NeelyCam
            • 9 years ago

            You’re all sorts of evil

            • Meadows
            • 9 years ago

            You’ll be one of the first to know as soon as I start giving a damn.

            • HisDivineShadow
            • 9 years ago

            The sheer number of people who say the same on Newegg, Amazon and the like suggest you are not alone.

    • dpaus
    • 9 years ago

    I think we all saw this coming…

Pin It on Pinterest

Share This