Attention hard drive manufacturers! Most people believe that a kilobyte is 1,024 bytes!

Attention hard drive manufacturers! Most people believe that a kilobyte is 1,024 bytes!

Summary: Operating systems use a binary system to measure disk space, while hard drive manufacturers use a decimal system. The result - When you fit a 750GB drive to your PC and fire up your OS, you only see 698.5GB - a whopping 51.5GB short of what most people expect to see.

SHARE:
TOPICS: Legal
54

It seems that the hard drive maker Western Digital has caved in and will offer free software to some 1 million consumers in order to resolve a class-action lawsuit that alleged that hard drives sold stored less information than promised.  To resolve this class-action, Western Digital is to give away backup and recovery software to anyone who bought Western Digital hard drives between March 22, 2001, through February 15 2006 (if you bought a hard drive and want to claim what's due to you under this settlement, you must register your claim before July 16 at http://www.wdc.com/settlement).

The problem is one of standards that I've talked about for years and I think that Western Digital has been wise to avoid taking this issue to court.  Operating system makers like Microsoft and Apple use a binary system to measure kilobyte, megabytes and gigabytes.  Under this scheme, a kilobyte is made up of 1,024 (210) bytes, a megabyte is made up of 1,048,576 (220) bytes and a gigabyte made up of 1,073,741,824 (230) bytes.  Hard drive manufacturers on the other hand use the decimal system for calculating the number of bytes that go to make a kilobyte, megabyte and gigabyte - 1,000 (103), 1,000,000 (106) and 1,000,000,000 (109) respectively.  The upshot of all this is that for every gigabyte fitted as storage to a PC, when this is measured by Windows or Mac OS, the customer gets 74 megabytes less than they expect to see.  That doesn't sound like a lot, but if you scale it up, it becomes pretty noticeable.  For example, take the latest 750GB perpendicular drives.  When you buy one of these and hook it up to a PC and fire up your OS, you only see 698.5GB - that's a whopping 51.5GB short of what most people expect to see (that's equivalent to nearly 6 full-length DVD movies).

The confusion goes deeper than just hard drives.  While CD capacities are given in binary units, DVDs are measured in decimal units.

Now you might be wondering how come this is legal and how come there hasn't been a lawsuit long before this one.  Well, the problem is that the trick could well be legal because the prefix "kilo" is a recognized prefix which is shorthand for 103 or 1,000 under the SI standard.  Under the same standard "mega" is short for 106 or 1,000,000 and "giga" is short for 109 or 1,000,000,000.  The prefix for binary representations are different, where "kibi" represents 210, "mebi" represents 220 and "gibi" which represents 230.

So is this a case of conforming to standards (the SI standards) or is it just a trick?  The kilobyte  = 1,000 bytes dates back to serially accessed data (such as punch cards) and the first hard drives used the same format.  However, times have moved on a lot since punch cards and I feel it’s got to the point where is starts to feel more and more like a trick so maybe it's high time media manufacturers accepted that, in computing circles, the binary system has become the recognized standard and start labeling drives based on what the end user will see.  If they are concerned about the perceived drop in capacity then why not adopt a dual system - such as 750GB/698.5GiB. Otherwise I think more and more people are going to notice what they're missing and more lawsuits will follow.

Topic: Legal

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

54 comments
Log in or register to join the discussion
  • not a trick; it's the "marketing gigabyte"

    This isn't new; power-of-ten sizing has been used for disk size since back to the 1980s. (Back farther than that in the mainframe world, where formatted capacity was also less than usable capacity for as long as there have been disk drives).

    What's new is that the difference between power-of-ten and power-of-two measurements has grown to be large enough to be a substantial hard-drive size of its own.

    Aside from technology issues, marketing will always want round numbers that are easy to advertise and easy to remember. And they will round up, not down.
    diane wilson
    • The difference will be big when 1TB disks become a reality

      Then the difference will be 10% ... that's a huge difference.
      Adrian Kingsley-Hughes
      • You're right, the error gets larger

        When you're comparing 1000 to 1024, it's a 2.4% difference. When comparing 1000000000000 to 1099511627776, the difference is 9.95%. Good point.
        georgeou
        • George investigate this:

          The same standards you speak of with hard drives is neerly the same as broadband speeds. I was crushed when I got cable as one of the first users that boasted about 1.5meg abit speed. then I goto download and get about 200 bits per second as time moves on and the speeds increas I am now on a 7meg dsl line and get about 4meg called them they say up to 7meg I wonder if they are perhaps involved in the same marketing tomfoolery as the HD manufactures and when I get a 120meg line in the future what will my real speed be 7megs?
          IceTheNet@...
  • Good point Adrian, I learned something new today.

    Of course, that 'most' is the problem: most people won't notice, but I'm glad you brought it up.
    BillyG_n_SC
  • How Much is that Kilobyte in the Corner?

    I find it ridiculous that people are wasting all this time, money and other resources over what is essentially a moot point. Yes, the difference is only 24 bytes per 1,000. Granted, that difference is multiplied 1,000-fold every time you expand to the next generation.

    <i>"OMG! My terabyte hard drive has only 976 gigabytes!"</i> Yeah, wait until you format that sucker ? are you going to sue <b>Microsoft</b> because of the overhead the File Allocation Table and other system files take up?

    I am convinced -- this is all the fabrication of a bunch of greedy shyster lawyers sucking up cash from a class-action suit. Remember the Iomega Zip "Click of Death" suit? I got <i>squat</i> from that. I do not remember what the attorneys walked away with, but it was substantial.

    That said, please grow up, get over it and move along. You are wasting my time, and my money, raising the price of my next hard drive purchase.

    <i>--Perfesser</i>
    perfesser_bear
    • A terabyte drive ...

      ... would have jsut over 900GB! :-)
      Adrian Kingsley-Hughes
      • would that include the fat

        and the master boot record or is that included in your estimation. because with a NTSF file system that takes the mysterious 7megs off the top I think you may be looking at having less than 900GBs for your actual storage. That is missing over a 100GB drive or in market terms about $100 of missing space. OUCH!
        IceTheNet@...
    • ZIP COD

      I remember what we got for that, I got a $12 (US) coupon toward another crappy Iomega product. Not only did I NOT get a new drive out of it, (they sold us defective drives, at the least they should have replaced the drives) but that $12 did not even cover the $20 I was out for the two disks the drive ate.

      Add to this the fact that because of how that whole thing was handled, there was no way I was going to buy another Iomega product, that coupon is still sitting in a folder with my reciepts all these years later.

      Oh, and standard in a class-action suit for the attorneys is one third of the total the defendant has to pay out (or the equivalent) so if you can find out how many people coupons were sent to, multiply that by $12 (US) then divide that by 2 and you'll have about the total the lawyers pocketed. That is only if they used that standard, if not I'd imagine the actual total can be found out, but at this point I don't think anyone actually cares . . . thing is, you're correct . . . the only people that actually win are the lawyers, the actual people hurt in these cases usually end up with squat.

      Mic
      micjustmic
  • The only winners are the lawyers, to the tune of $500,000...

    "San Francisco lawyers Adam Gutride and Seth Safier will receive attorneys? fees of up to $485,000 and expenses up to $15,000 for their work in the case, pending approval of the proposal. The pair has also filed a similar lawsuit against Seagate Technology."

    http://www.betanews.com/article/Western_Digital_Settles_Capacity_Suit/1151510648

    What a shame. Ok, so maybe a few customers got a few less gigs than they thought they should. Maybe those customers do deserve a small voucher. But the only ones who are *really* benefitting from this are the lawyers. What a crock.
    PB_z
    • Yeah ...

      This is why I don't like lawsuits ... they are a scam. Shame companies can't behave themselves.
      Adrian Kingsley-Hughes
    • Absolutely agree

      I absolutely agree.

      To get yourself added to the list of users on the suit you have to get the serial number off the hard drive.

      Now the real question is if its worth me taking the 30-60 minutes to take apart my system to get the number and then put it back and hope nothing goes fubar because I was messing with all of the cabling.
      Silent Observer
  • Bigger problem is the tape manufacturers

    I've noticed in Europe, tape manufacturers market their tapes at true capacity. In the US, they market it as 2x capacity. That's a much bigger problem for consumers since photos and DVDs don't compress.
    georgeou
    • isn't that one heck of a wrinkle!

      think of it this way...
      on one hand the HD manufacturers are using the kilo reference to suggest that the hard drive actually has more capacity than it does instead of how much actual space the drive has once formatted by the OS.

      on the other hand the tape drive makers use a number for capacity AFTER the tape has been "formatted" and after "compression" is taken into account, a function that is primarily software related, to suggest again that the tape stores more capacity than it does.

      is it just me or does it seem like they're playing a lil fast and lose there?
      here's the drive, it has this capacity, but really only this when you format it.
      here's a tape, it really only has this capacity, but it has THIS much when you format and use compression.

      O_o

      Valis
      Valis Keogh
    • They do that? That IS a ripoff!

      They market based on [i]compressed[/i] capacity?

      That's a big ripoff - you never really know how much something is really going to compress! Some things compress well, and some barely compress at all. That's a pretty big risk, and a horrible ripoff for people storing a lot of stuff that doesn't compress very well.
      CobraA1
      • It's true

        It's been that way for almost 8~10 years as far as I can remember. Whenever you read the label of the tape, you'll notice a notice at the bottom written in very fine print: [i]*2:1 compression ratio.[/i]
        in-DUH-vidual
        • re: It's true

          Oh, they've been doing that for at least 15 years that I've seen. Probably longer.
          johnay
          • It goes back to the beginning of time (hard drive wise)

            ... only that it's now starting to become obvious to buyers.

            The question is always worse when it coems from someone who's bought a new PC - they immediatly start to accuse the vendor of swindling them (same with shared graphics memory).
            Adrian Kingsley-Hughes
  • This is OLD news

    This has been a problem since Windows 95. Maybe even earlier. You just noticed this?!?!
    CobraA1
    • Western Digital settlement is not old

      Did you not notice that?
      Adrian Kingsley-Hughes