I’ve just started noticing websites (maybe it’s just Wikipedia?) referring to computer storage capacity in GiB and MiB. For example, a DVD holds 4.7 GiB of data. Today, I finally got around to investigating what those abbreviations mean. They’re gibibytes and mebibytes, of course!

Some time after 3.5-in. floppies existed and before I ever heard the word “gigabyte”, I became aware of the awkward inaccuracy of the whole kilobyte and megabyte counting scheme, something that most web surfers are oblivious to: that a kilobyte is 1,024 bytes, not 1,000 bytes!

It’s not really that big of a deal. So what if 5,678,901 bytes is really 5.5 MB instead of 5.6 MB? It either fits on the disk or doesn’t.

But this is where the mebibytes and gibibytes come in. According to Wikipedia, this, so called, binary prefixing scheme has been in use in the geekiest computer circles since 1964. Did they even have a thousand bytes of digital storage in 1964? It works like this:

Because the metric system prefixes require the use of base-10 multiples, we define a megabyte as what everyone already thinks it is: 1,000 kilobytes (1,000,000 bytes). Then, we create this alternative base-2 prefixing scheme that accurately describes the quantities in the way that computers prefer to think about them: 1 MiB = 1,024 KiB = 1,048,576 bytes. This way, we can avoid any “why are my 5,678,901 bytes only 5.5 MB?” confusion.

I love it. This naming scheme appeals to the geeky programmer/scientist in me. I’m going to try to adopt this prefixing system when writing. I’m not sure I can bring myself to actually say the word “gibibyte” aloud, though.