How many Gb do you lose?

SideWinder · 18 Mar 2007 at 16:25

How many Gb on average do you "lose" when you install HDDs? Are we talking a few Gb or over 10?

And kind of related, could you still say you had 1Tb of storage if you have 2x500Gb or would you personally include the lost Gb?

mosfet · 18 Mar 2007 at 16:32

This has been discussed a few times before on here, lots of misunderstanding around the subject:

http://forums.overclockers.co.uk/showthread.php?t=17649350

You don't lose any space. 2x500GB drives will give you 1000GB or 1TB total space. What it won't give you is 1000GiB or 1TiB.

Take a look at http://en.wikipedia.org/wiki/Binary_prefix if you want to know more.

SideWinder · 18 Mar 2007 at 16:37

Hmm. Reason i ask, my friend has "lost" around 11Gb on a brand new computer. And i was always under the impression you lose a bit...my C drive is 27.5Gb and my D drive is 83Gb so...

I'm not interested in what they call 1TB hard drives, as they'll only have a capacity of 931GB.

'Only?', I hear you say. Well I'm a semi-obsessive sort, and if I want a terabyte then I want an ACTUAL terabyte, so that Windows will report the capacity as such.

mosfet · 18 Mar 2007 at 16:40

'Lost' from what?

From the stated capacity of the drive to the formatted capacity in windows? Or from what the computer manufacturer stated to the free space capacity of the drive?

As for your quote, it would be correct if it read: "...as they'll only have a capacity of 931GiB."

I suggest you thoroughly read the links I posted, it will probably explain what you need to know.

SideWinder · 18 Mar 2007 at 16:43

I'll read the thread more closely. Just saw Page 2 and it has a LOT of writing.

james.miller · 18 Mar 2007 at 16:46

windows reckons 1gb = 1024mb. hard drive manufactures think 1gb = 1000mb. thats why windows reports less space on the drive, because there are less 1024's than there are 1000's in any given number.

easy as that. You dont 'loose' anything at all, besides a small amount that goes to the fat table ect

SideWinder · 18 Mar 2007 at 16:48

james.miller said:
windows reckons 1gb = 1024mb. hard drive manufactures think 1gb = 1000mb. thats why windows reports less space on the drive, because there are less 1024's than there are 1000's in any given number.

easy as that. You dont 'loose' anything at all, besides a small amount that goes to the fat table ect

Thankyou.

So if i had 2x500Gb i could "say" i have a Tb of storage but in actual fact i can't use every single Gb due to the Windows and manufactures differences?

mosfet · 18 Mar 2007 at 16:52

You would indeed own 1000GB of storage space, or 1TB. This is equal to about 931GiB of space, which is what Windows defines.

james.miller · 18 Mar 2007 at 16:53

bingo. it all comes down to the confusion between GB and GiB

SideWinder · 18 Mar 2007 at 16:54

Thankyou.

Sorry for asking such a silly question but it's the only way i'll learn. <3

PhillyDee · 18 Mar 2007 at 17:55

Its not just confusion, its done intentionally by the disk manufacturers. Means they can get away with giving you less space (than you think you are getting). It only really becomes apparent if you see it written as bytes of free space.

mosfet · 18 Mar 2007 at 19:58

PhillyDee said:
Its not just confusion, its done intentionally by the disk manufacturers. Means they can get away with giving you less space (than you think you are getting). It only really becomes apparent if you see it written as bytes of free space.

Wrong. Read the links posted above.

quantumisation · 19 Mar 2007 at 00:23

they really should change this, id like my computer to look all tidy with round numbers and not 465 and 279 etc. Damn hard drive manufacturers. Yes i could partition to make even numbers but thats hassle and youd lose space, whats next defining terabytes as 100 gigabytes? but hey if thats how they define it then thats cool

mosfet · 19 Mar 2007 at 00:47

*sigh*

Hard-disk drives store data in a serial manner and so takes the counting base of time - 10. Whilst varying between a few manufacturers (Iomega and Quantum), it has essentially been like this since the advent of the technology in the '50s. HDD manufacturers did not define the system to rip people off, they used the SI system that everyone else in the world of science uses.

CPUs and other electronics work to base 2 for a whole variety of reasons.

This is where the confusion arises. To stop the confusion between bases the IEEE and IEC defined a binary prefix system. Windows and Mac OS do not yet implement it.

Enlighten yourself:
http://en.wikipedia.org/wiki/Binary_prefix

ByteJuggler · 19 Mar 2007 at 12:36

Yes, it's simply a case of conflating different measures/measurement standards. If anything, Windows should probably rather start using the SI measurements given that its user interfaces is meant to be "human oriented" rather than computer oriented (in which case using base 2 measures are probably fine.) Ideally though, if I was writing Windows, I'd enhance it to show both measures with some clarifying hints or so, or allow you to switch between which measure was used with e.g. a tick-box or so. Then there'd be no confusion. But nooooooo... :rolleyes:

Corasik · 20 Mar 2007 at 11:18

Binary powers make more sence for computers though, as modern computers are binary machines. Switching windows to SI would be just as confusing, as then you would have to say 1GiB of memory was 1.074GB memory.

Regardless of what Symbol, and word they use, GiB, GB, TiB, TB. Gigabyte, Gibibyte.... The computer industry really should pull together and use the same standards all round.

The average joe is never going to understand that a computer has 2GiB memory, and 250GB of hard disk, with a 700MiB CDRom, and a 4.7GB DVD, and for flashing bios he can use a 720KiB floppy, or a 1.44MB floppy.

It's a mess for the consumer, and because modern computers are inherently binary devices (Unless that changes memory will remain sized by binary powers) everyone should use the binary version. Hard drives, Ram, Optical media.

So, this whole MiB, GiB, TiB standard was made in 1999 and the standards boards seem to like it, but its hardly in 'general' use. Sounds more like Men in Black that a computer term to me

but thats beside the point.

Cant see anyone going into the shops to ask for 1Mebibyte of ram, and I guess you would have to shorten Mebi to Meb, So no more Gig's, or Meg's. Perhaps its just me, but Mebi, and Gibi and Tebi etc just dont 'sound' right when spoken. Probably because its too many 'b's when you say it out full. Kibibytes etc.

Guess im just too old, after 25 years of calling 1024 bytes a kilobyte I dont think im likely to start calling it a kibibyte because some pencil pushers in the 'standards commities' think its a good idea.

Corasik · 20 Mar 2007 at 11:41

ByteJuggler said:
Yes, it's simply a case of conflating different measures/measurement standards. If anything, Windows should probably rather start using the SI measurements given that its user interfaces is meant to be "human oriented" rather than computer oriented (in which case using base 2 measures are probably fine.) Ideally though, if I was writing Windows, I'd enhance it to show both measures with some clarifying hints or so, or allow you to switch between which measure was used with e.g. a tick-box or so. Then there'd be no confusion. But nooooooo...

Why show anything other than the binary power version, inside the OS at all, hard drives are sectored up in 512k blocks (NTFS clusters are groups of sectors, so multiples of 512k). So while a single file may be any size, the space it takes on disk will always be a multiple of 512, or more accuratly a multiple of the formatted cluster size (normally 4096K). Drives are formatted this way, because computers CPU's are binary, so everything they do is based on binary powers. Our computers deal with 8, 16, 32, and 64bit wide busses and numbers, and thats how computer memory works. Its only logical that the disk storage is divided into blocks that make the most efficient use of the processors binary nature.

ByteJuggler · 20 Mar 2007 at 12:29

Corasik said:
Why show anything other than the binary power version, inside the OS at all, hard drives are sectored up in 512k blocks (NTFS clusters are groups of sectors, so multiples of 512k). So while a single file may be any size, the space it takes on disk will always be a multiple of 512, or more accuratly a multiple of the formatted cluster size (normally 4096K). Drives are formatted this way, because computers CPU's are binary, so everything they do is based on binary powers. Our computers deal with 8, 16, 32, and 64bit wide busses and numbers, and thats how computer memory works. Its only logical that the disk storage is divided into blocks that make the most efficient use of the processors binary nature.

WE know that, bute we're techies. Users are not techies. I used to share your viewpoint at one point, but became convinced that one standard is just as good as another, but, since SI is what is used for everything else, I think it should also be the preferred choice for computers, unless the application has a particular use for it and it makes sense in the context to use base 2. Why expose general users to any powers of 2 at all? It only breeds confusion as this type of thread recurring has made very clear. Users are used to base 10, thus that is what UI's should use as a preference, unless it's made abundantly clear that this is not the case. If it's a technical application where there's some use to be had from listing/using power of 2 based numbers then fine, use it, but for general purpose applications, sticking to base 10 is better IMO. Computers after all are not an end unto themselves. For general end-users, they should be made to be as user friendly as possible. Consequently using SI for measurements in general is one way of making it. Pedantically sticking to base 2 when the only motivating factor is that that is how it works under the hood is not good enough IMO.

At the very least, Windows should preferably use the correct measurement indicator, which it does not:

1GB = 1Gigabyte = 10^9 bytes
1GiB = 1 Gibibyte = 2^30 bytes

If they simply inserted an "i" into their size displays, everything would at least be correct according to the official definitions of the measures, and confusion should then be reduced.

mosfet · 20 Mar 2007 at 12:41

Corasik said:
It's a mess for the consumer, and because modern computers are inherently binary devices (Unless that changes memory will remain sized by binary powers) everyone should use the binary version. Hard drives, Ram, Optical media.

Regardless of this the two bases are required - transmission speeds are always counted in a base-10 system because that's what we measure time in. Even if the decimal-base storage systems such as FDDs, HDDs and DVDs become redundant, the transmission rates of busses, the speed of the processor etc are all in Megabytes/Sec or Megahertz. If you define Megabyte as 1024 Kilobytes, you can now argue that you're being ripped off by the processor manufacturer for selling you a processor that clocks at 2000MHz instead of 2048MHz.

Guess im just too old, after 25 years of calling 1024 bytes a kilobyte I dont think im likely to start calling it a kibibyte because some pencil pushers in the 'standards commities' think its a good idea.

I wouldn't call the IEEE and IEC 'pencil pushers'. They define and help develope virtually every computing standard. Without them there would be very little interoperation between competing systems.

Why show anything other than the binary power version, inside the OS at all, hard drives are sectored up in 512k blocks (NTFS clusters are groups of sectors, so multiples of 512k). So while a single file may be any size, the space it takes on disk will always be a multiple of 512, or more accuratly a multiple of the formatted cluster size (normally 4096K). Drives are formatted this way, because computers CPUs are binary, so everything they do is based on binary powers. Our computers deal with 8, 16, 32, and 64bit wide busses and numbers, and thats how computer memory works. Its only logical that the disk storage is divided into blocks that make the most efficient use of the processors binary nature.

I agree that operating systems should use binary powers for displaying file information, but transmission rates should stay in a decimal base.

If Windows and MacOS start using the 2 systems as they are supposed to be used then the general computer user might start to realise why there are two systems and how they're not simply being 'ripped-off' by hard-drive manufacturers.

Fr0dders · 20 Mar 2007 at 13:08

Corasik said:
Guess im just too old, after 25 years of calling 1024 bytes a kilobyte I dont think im likely to start calling it a kibibyte because some pencil pushers in the 'standards commities' think its a good idea.

its not a question of it being a good idea, its been wrong. And just because you've been calling it by the wrong name for 25 years, doesnt change anything

the SI system of units is base 10. So 1 kilometre is 1000 metres. 1 Megametre is 1,000,000 metres, etc..

1 kilobyte is 1000 bytes. in exactly the same way that 1 kilolitre is 1000 litres etc..

Computing should never have used the SI unit names for its base2 system. It only confuses consumers, and windows should at least correctly name the ammount so consumers dont get confused.

clock cycles are counted in Base10, so not everything in computer is in the powers of 2.