A Petabyte in Gigabytes: The Ultimate Guide to Units Of Information

Back in the late 80s/early 90s, my dad acquired a PC which ran on the MS-DOS operating system. The darn thing was so expensive that I recall he had to pay for it in many multiple installments.

The machine had around 10 or 20 Mega Something of hard-drive (HD) space and 2 Mega Something of RAM, which went from being loads of space and loads of memory to being insufficient really quickly. (The reason for Mega Something should become clear as you read).

You probably guessed why it became insufficient: it was all due to games!

To get games, I would either need to beg my parents to buy magazines that came with floppy disks or I would get them by copying from friends that also got them the same way.

Mind that I had no concept of what piracy was and I had just started to learn English, so there was loads of trial and error to figure out how things worked.

I recall getting angry when the games were hard to play (needing good English or being too complex to figure out) or when there were more tools than games on the floppies, which had 1.44 Mega Something.

Little did I know then that those tools would eventually help.

Led by the need to extend the little amount of HD space I had, I began teaching myself how to partition files into chunks so they would fit in floppy disks so I could later uninstall something and re-install the large files back on disk.

Although I had to do that often, I always took extra care to ensure everything was always working properly in case someone else needed to use the PC.

As time passed, the rest of the family was not using the computer much, probably because I was always on it, so I started taking some liberties.

Without fully understanding the risks, I eventually learned about executing a command called DoubleSpace (later renamed Drivespace), whose dark magic made the HD suddenly have double the capacity (20 or 40 Mega Something).

Oh, the Joy!!! I could not believe it! I think I’ve tried it a few times just to make sure my eyes were not deceiving me! And I recall screaming out of happiness. Everything was working fine and the HD had so much more capacity!

But that didn't last. Eventually, the system began to have issues, the hard disk started to have bad sectors and there were other errors that came out of nowhere.

With no one to ask for help, and as the magazines didn’t point to the command as the culprit, I had to spend several days in front of the computer trying out tools that came with the magazines to recover from Microsoft’s buggy Software to fix the HD.

I also had no concept of what compression was, until I read about it in a magazine that came with this cool new utility that made life so much easier: Zip! It reduced my information transfer efforts, but it came a bit too late to save the HD.

Those were good days spent on problem-solving, learning, and gaming. Man, did I squeeze the life out of that little hard drive!

When I learned it, we had:

There were bigger units in 1991, but I didn't care much about:

A Bit was, and is, a 0 or a 1 and a Byte has 8 bits. Historically a Byte would have different values than 8, but I’ve learned it as 8 (dodged that one somewhat, perhaps a topic of interest for another day).

And an 8 Bits Byte is called an octet. A term that unambiguously describes the number of bits in a Byte.

A Kilobyte was 1024 bytes and a Megabyte was HUGE, holding 1024 Kilobytes.

Life was simple, I had no network, all I cared about was having a way to express raw capacity (the amount of information a device can hold). Or did I? You’ll find out.

As all my time was spent at the computer, and despite being very young, I became the local boffin (a person with knowledge or a skill considered to be complex) to those that knew me (not that many, I was still a kid).

“Why is it 1024 and not 1000 Mega Something?” This was a question I would need to respond to often and to which I would happily respond, expert style: “I don’t know, it’s just like that!”. (Boffin no more, right?)

The fact is, as time passes, this question gets trickier to answer. And to make matters more interesting, the way I learned it is no longer (exactly) valid.

Why? Well, we have more than one system. One is based on the power of 10 while the other is based on the power of 2.

Which one we use should depend on what we need to measure and according to who we want to measure, as some multiples of the units are formed with either SI prefixes (power of 10) while others use IEC binary prefixes (power of 2).

Let’s look at the following table that splits decimals (power of 10, 0 to 9) with binary (power of 2, 0 and 1):

As you can observe in the above table, the Bit and the Byte have the same value, but it changes from there onwards with substantial differences between the values, which identifies its relevance.

The larger the unit, the larger the difference, and that’s what makes it important.

But it can get confusing.

For example, a MAC can be advertised in the Apple store as having a HD with 1 Terabyte. If you would buy one, then you could go to “System Preferences” > “General” > “About” in Ventura, where you will be able to confirm that it does have the Terabyte.

However, if you would enquire in the command line using the “df -h” command, then you would get the response in Gibibytes, precisely 932 Gi (not GiB, which it abbreviates; perhaps the author will fix that in the next update).

So hold your anger; now you know Apple measures in binary and did not steal almost 10% of the storage capacity you paid for.

This happens in a lot of other circumstances too and is not restricted to Apple. So take care when you are buying something or need to be precise, particularly around old Software that may be referring to Mebibytes but naming them Megabytes. As in those, the Megabyte may still be equivalent to 1024 Kilobytes and not 1,000.

You may be asking: Why do we need this? To me, it has to do with human needs, in this case, our need for the familiar and for consistency. In the case of SI prefixes, they follow the same units as the metre (British English spelling).

If 1 Litre equals 1000 Millilitres and 1 Ton equals 1000 Kilos and a Metre equals 1000 Millimetres, why would a Megabyte be 1024?

And this was the feedback I used to get when young, from adults that asked: “Why is it 1024?”

In other news, since November 2022, Ronnabytes and Quettabytes have been adopted by the International Bureau of Weights and Measures and were added to the International System of Units (the SI I’ve mentioned before), so now we have two more to know.

Yup, they have a lot of Bits!

Let’s first observe how many Bits some of the units refer to:

However, the precision (and confusion) does not end here.

We also have the Megabit and Mebibit. No, I’m not making those up.

When you want to refer to the amount of data transferred in a network (or telecommunication systems) as to figure out how fast it is, then you should use Megabits or Mebibits, which consider time.

To transfer 1 Megabyte in a second, you would need a network connection with a transfer rate of 8 Megabits per second.

Example 1: 1 Megabyte (8,000,000 Bits) divided by 1 Megabit (1,000,000 Bits) results in (theoretically) 8 seconds of transfer speed.

Example 2: 1 Mebibyte (8,388,608 Bits) divided by 1 Mebibit (1,048,576 Bits) results in the same 8 seconds.

The Mebibit, as you probably guessed by now, measures in powers of 2.

There you have it! You should now be able to scroll up and replace the Mega Somethings with the desired precision. I hope you get them right.