(1 MB == 1024KB) or (1MB == 1000KB) ?

Hope this is the appropriate forum.

Many years ago I read this:

1 Kb is 1024 bytes

and

1 Mb is 1000 Kb ( NOT, 1024 Kb ).


As a confirmation to that,
the floppy images size for installing freeBSD is 1474560 bytes.

This is obtained with 1000 * 1024 * 1.44

( A floppy is 1.44Mb )

Any comment about this dilemma ?!?

http://scienceworld.wolfram.com/physics/Megabyte.html

1 KB = 1024 Bytes
1 MB = 1024 KB = 1,048,576 Bytes

You can format a 1.44 MB disk to different capacities. The floppy in my Windows machine can store 1,457, 664 = 1.38 MB. On a Linux box you can get the following (http://floppyutil.sourceforge.net/man/floppy.html):

Thank you for the links.

I always thought that about "memory addressing" and "memory capacity" me too.

As an example "my" microcontroller (an 16-bit hitachi ) can address up to 1MB,
( address bus from A0 to A19 ).

But I thought that when talking about disk storage informations,
it was different.

I found this link that maybe could help to distinguish:

http://www.romulus2.com/articles/gu...bitsbytes.shtml

There I read that:



The last note maybe could explain why I read that in the past.

Perhaps this site is interesting, too.

Everything related to computing is in base 2, i.e. 2^x, this includes disk sizes.

Unfortunatly, most consumers don't understand binary numbers and so retailers have a nasty habit of runding these figures off to something more decimal.

This has lead to a bluring of what is and is not correct but if you want to be accurate

1 byte = 8 bits
------------------
1 Kb = 1024 bytes
1Mb = 1024Kb
1Gb = 1024Mb
etc...

Don't always believe what you read!

if you use the 1000, a harddisk you buy will look bigger than it actually is... marketing strategy.

some years ago i also read about MB (MegaByte=1024*1024*1024) and mB (million Byte=1000*1000*1000). but never heard of this again in the present...

Some people would beg to differ. In KILObyte the prefix kilo is used. This means 1000. So people can say that 1000 kilobytes are in a meg... the thing is, it doesn't matter what the prefix means, the megabyte can hold 1024 kilobytes. I believe that 1024 kilobytes is a meg

look at these:

2^10
2^20
2^30
2^40
2^50

See a pattern

[eko]

The well-known pattern is ai = 2 ^^ (i * 10), i being the index of a prefix in {kilo; mega; giga; tera; exo; ...} - so what?

Matt Light:
This is a bit inconsistent, isn't it? If you assume that kilo is strictly 10 ^^ 3, then you also have to equal mega to 10 ^^ 6 or use the approximations with powers of two for both, otherwise the prefix system becomes pointless. What if someone decides that giga should be 1048 * mega? It would fit your pattern, but doesn't make much sense for computers neither for humans.

Given that we have all been brought up on the SI system, it should be clear that the prefixes kilo, mega and giga have the following meanings:

kilo = 10^3
mega = 10^6
giga = 10^9.
tera = 10^12

Clearly, the idea that
kilo = 2^10
mega = 2^20
giga = 2^30
etc...

is a convenient fiction, created by computer scientists who found that powers of 2 suited them better. It is, however, a non-SI usage of those prefixes. So, at the very least,

1 megabyte = 10^6 bytes.

Whether MB stands for Megabyte or something else remains to be seen.

Remember, all you powers-of-two fans, that tradition does not a standard make.

If you really want, you can define your basic unit of time to be 1/18.25 SI seconds. You can even call this new unit the second. But a second has a standard definition, and so does a {kilo || mega || giga} of anything.

With kilo the SI convention is to use the small k. So I make the distinction:
1 k == 10^^3
1 K == 2^^10
this sadly can not be extended for the higher prefixes at mega, giga and tera by default use capitals (plus the small m is allready used by milli).
So basically the best is to use scientific notation. This also avoids the pitfall
Europe:

(oops I pressed return once too often so this is the last post continued

europe: 1 billion == 1 million million
USA: 1 billion == 1 thousand million

if i remember correctly: the SI is defined for a base of 10 always. not 2 and not 16. CPUs never work on a base of 10.

and: aren´t people in the ´states and the uk still using miles and ounces? - no rule without exceptions.
(correct me if i am wrong )

i prefer the 2^10 way when talking about bits´n´bytes. this is the way i remember it has always been...

And as a side-effect it makes calculations sooo much easier (if you ever did assembler / low-level c programming)

SI is designed for metric units. Clearly metric units are unsuitable for computer storage ability (apart from tape and earlier stuff), since the design of a common computer operates with power-of-two values.

Of course you could force everyone to use the decimal stuff, but then you'd also have to use it in programming, OS (assembler) programming and hardware design. Very inconvenient, given the certainty that the new stuff wouldn't be compatible to the old stuff.


Yes. Going through thirty lines of rol, ror, shl, shr etc. would be even worse if you had to convert to decimal and back all the time.

letz make a parallel-processor that does like 2^3+2 bits at once, anyone interested??

me neither.... humans are not compatible, cpus are