I have a number of thumb drives and hard drives. One hard drive I bought was said to be 2TB but its actual capacity is about 1.8 TB. A couple of thumb drives I got say they are 16 GB but on the back say:

1MB = 1,000,000 bytes /
1GB = 1,000,000,000 bytes

This definition is the reason why I've accepted the false advertising on hard drive space, even though conventionally:

1KB = 1024 bytes
1MB = 1024x1024 bytes
1GB = 1024x1024x1024 bytes

and not

1KB = 1000 bytes
1MB = 1000x1000 bytes
1GB = 1000x1000x1000 bytes

As we get to higher levels the amount of space that is lost by this rounding down increases.

But isn't this still false advertising? E.g., in Australia could I get a refund as I was misled in the believe that a thumb/hard drive had the higher capacity?

  • ... Have you bothered to read the letters?
    – Zizouz212
    Mar 22, 2016 at 13:14
  • That's the actual value. And they are not lying - they even gave you a definition of the term when in doubt!
    – Zizouz212
    Mar 22, 2016 at 13:15
  • They can do that because the actual definition is 1GB = one billion bytes, so it’s absolutely correct.
    – gnasher729
    Jan 30, 2020 at 10:35
  • Nobody ever complains that a kilogram of flour is not 1024 grams, isn't it?
    – lalala
    Jan 20, 2022 at 7:23

3 Answers 3


It's not the manufacturers who are wrong. Your definitions of KB, MB, and GB are incorrect. See, for example, NIST. The numbers they are using are not "rounded down," they are the proper standardized definitions of those terms. 1024 bytes is properly termed a kibibyte (or KiB), not a kilobyte. 1024 kibibytes is a mebibyte, MiB. 1024 mebibytes is a gibibyte, GiB. These are defined in an international standard, IEC 80000-13 (part of the ISO series of standards defining units); "kilobyte" refers to 1000 bytes, "megabyte" to 1000000, and "gigabyte" to 1000000000. It's not lying to use units correctly, and that's what the manufacturers are doing.

  • So that means Windows is lying as i have my 16 GB thumbdrive in and it says the capacity is 15,818,031,104 Bytes but on the side it says 14.7 GB, not 14.7 GiB and 14.7 is the approximate value of (((15,818,031,104)/1024)/1024)/1024.
    – Memor-X
    Mar 22, 2016 at 13:24
  • 1
    @Memor-X Windows is incorrect with its definition of GB. That's not on drivemakers.
    – cpast
    Mar 22, 2016 at 13:30
  • 4
    @Memor-X I would say "controversial" or perhaps "old fashioned" rather than "incorrect." The use of the SI prefixes to mean multiples of 1024 predates the development of KiB and the like by some decades.
    – phoog
    Mar 22, 2016 at 13:50
  • 7
    Whether the definition is incorrect is a red herring. Each gives you a way of knowing how many actual bytes available. The drive manufacturer by their note on the package staying which definitions they are using, and the OS by giving you more detailed info down to the byte if you look for it.
    – user3851
    Mar 22, 2016 at 13:53
  • 2
    Originally, in computer mathematics, a KB was 1024 since computer internaly do not use decimal numbers. Manufacturers started cheating almost from day one when they started selling hard disks giving the impression that they are offering more than they are giving, knowing fully that then (in the 1980's) buyer were calulating their needs based on 1024 per KB. In 1993, insteading of forcing the cheaters to give an honest value, they expected the Computer Mathematics to change every use of KB since the 1960's to KiB This did not happen. Jan 30, 2020 at 8:16

As well as the accepted answer, don't forget that the reason your drive is reporting free space below the capacity, is dependent on what disk format you have used (or was done for you). Even a basic boot sector, partition map and root directory may take away hundreds of MB on a sufficiently large capacity drive.

There will be a difference in the 'free' space if you format a drive using FAT32 or NTFS or ZFS or [insert other filing system here].

Back in the last days of MS-DOS, 3.5" installation disks used a special format that would actually store 1.6MB instead of the more normal 1.44MB that you got from FAT12/16, by reducing the size of the root directory (amongst other tricks).


Almost all of the flash memory chips used in thumb drives hold a power-of-two number of 528-byte pages; in each page, 512 bytes are used to hold user data while the remainder are used for error-correction and bookkeeping. Some pages will be marked as defective by the manufacturer, and others are used as "slack space". The way flash memory chips work, doesn't allow the data in individual pages to be directly replaced. Instead, it's necessary to perform updates by writing new pages without immediately erasing old ones, and periodically relocate useful pages so as to allow for block-erase operations to erase larger chunks of old pages as a group. In many cases, a willingness to set aside more slack space can improve performance, and a desire to maximize storage efficiency will impede performance. A manufacturer could make a "two gigabyte" thumb drive that holds 2,000,000,000 bytes of user data just as cheaply as one that holds about 1,800,000,000 bytes of user data, but at the cost of performance and reliability. Having better performance is more useful.

Of course, this is assuming the manufacturer and seller aren't crooks. A lot of "32GiB" thumb drives are actually much smaller drives whose formatting is corrupted to make them appear bigger, but which actually means they can't hold anything reliably.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .