LEt's round 1024 down to 1000
Since hard disk manufacturers don't count the last 24 bytes in a kilobyte, perhaps they've only sold 976,562,500 drives instead of 1,000,000,000.
Seagate is celebrating the shipment of its one billionth disk drive after 29 years in biz. The storage giant reckons it will reach its second billion in less than five-years' time. Seagate said it's shipped the equivalent of 79 million terabytes of storage since the company made its first hard drive in 1979. The ST506 hard …
Is that a billion drives in standard counting, or in hard drive manufacturer counting, similar to how they calculate drive capacity? If the latter, I suppose that would make the actual number of physical drives they've managed to crank out somewhere nearer the neighborhood of twelve.
When you buy a dozen eggs, you don't get 16. When you buy 100 CDs, you don't get 128. When you're going 30km/hour, you're not going 32,768 meters per hour.
The only reason "kilobyte", "megabyte" and "gigabyte" have been used to represent 1,024, 1,048,576, and 1,073,741,824 bytes respectively is because it was and is illogical and impractical to package RAM in quantities of 1,000, 1,000,000, and 1,000,000,000 bytes. It was easier to refer to 1,024 bytes as a kilobyte rather than 1.024 kilobytes. THAT's where the inaccuracy is. Hard drive manufacturers (bless their kind souls) have been maintaining the purity of the decimal prefixes all this time, bearing the full brunt of your petty wrath.
The "U.S. Billion", when applied to such things as memory or hard drives, follows the SI prefixes (http://physics.nist.gov/cuu/Units/binary.html). It's been nearly a decade since these standards were adopted. If the drive makers can't get their act together and use the right prefixes, then don't blame the U.S.
Steven, they haven't been maintaing the purity of the decimal prefixes all this time. There was a time when they used powers of two. Up until drives hit about 200MB. In one week, I bought a WD Caviar drive with geometry of 987x12x35 which was labelled as 200MB and then another instance of the same model drive (WD1200 or something) days later with the same geometry of 987x12x35 and it was labelled as 212MB (it's 212244480 bytes).
That's the week they changed their behaviour from one that make sense to computer scientists to the marketing-prefered "our drives are bigger now" model.
The documentation for the drive claimed they were doing it for consistency because DOS' CHKDSK command quoted megabytes as being 1,000,000 bytes. Which is FALSE. CHKDSK always quoted bytes, with commas separating the thousands. It never claimed that the second comma from the right denoted a megabyte.
...about my wondering when a single drive would have the same capacity as their combined sales to date, but then I saw the first three comments.
That's well and good, but I would like to point out, for the record, that if anyone so much as *thinks* about advocating that we all go around saying "mebibytes" and "kibibytes", I swear on all that is holy that I will hunt him down and murder him with a plastic spork.
That is all.
A few comments have hit on it, but the top 3 just scream too much ignorance to be left alone.
MB = 1,000,000 bytes
MiB = 2^20 Bytes (1,048,576 bytes)
GB = 1,000,000,000 bytes
GiB = 2^30 bytes (1,073,741,824 bytes)
So when you buy a 500GB drive you get 500,000,000,000 bytes of storage. Not 500*2^30 bytes of storage.
We all know that 1024 does not equal 1000, don't make your stupidity known by ignoring long existing standards. Try following Acme Fixer's link, and you'll see that this standard was adopted about 10 years ago for dealing with binary quantities. I will agree with Wiernicki that nobody should adopt these in speech, but but at least be aware of the facts.
"It's better to remain silent and appear foolish than to open your mouth and remove all doubt" - take note
Devil bill because too much of that 79TB is housing bill's handy work (rather, somebody else's handy work that bill bought and subsequently ruined, quite infectiously)
"Not only did I hear a similar story which relates to Moore's Law, but apparently if the cars had reduced in size in a similar fashion to storage devices, your Rolls Royce would be the size of a matchbox."
That is quite good - considering that a matchbox cost less than two euros... except that it obviously does not have a working engine etc... but this could not possibly matter since you could not drive it anyway.. on the other hand matchbox cars are apparantly still popular considering that both of my daughters love to play with them most every day.
Paris - because I bet she has a collection of hot wheels in the closet...
I find it quite curious to see just how the perception of size has changed in the eye of the market. One is arguing about the "purity" of decimal, the other is bringing forth a difference between MB and MiB (a fan of Will Smith, perchance ?).
That's all nice and cute, but the facts are these :
At the beginning of the hard disk, manufacturers did indeed represent size in its proper power-of-2 base. 1024 is a kilobyte whether you like it or not, and a megabyte (MB, not MiB, which is term that came into being around Y2K and was probably invented to placate those who do not comprehend what MB actually stands for) is 1024 kilobytes (or KB, not KiB either).
One might argue that giving sizes in their proper base-2 format was confusing for the layman, but given the amount of hooplah that has been going on for the past eight years (and continues unabated) in the graphics arena around shaders, texels and megahertz, I seriously doubt the validity of that argument. Besides, the layman doesn't actually need to understand what the exact size of his storage is, all he needs to know is that he's getting more every year for less money.
Occam's razor states that the truth is much more simple : Seagate & Co discovered that they could twist the truth a bit and label their disk sizes in megabits or somesuch, thus implementing that wonderful "decimal purity" and, more importantly, cheating us out of an ever-growing proportion of what we should be getting.
And when you were buying the megabyte (so that there's no confusion) at over $10, the sensation of being cheated was a particularly expensive one.
Of course, given the price of storage today, I don't really mind any more. But I won't forget either.
Well said.
Of course, the real reason for having things like 1GB == 2^30 bytes is that it can be addressed by exactly 30 bits. Admitedly its more relevant for memory than disks (otherwise you'd find that when you install four 1GB memory sticks each holding 10^9 bytes that you've got annoying holes in the physical memory map). On the other hand, a sector is still 512 bytes (2^9) and file system blocks are an exact number of sectors (eg 2^12) for exactly the same reason: given an offset in bytes the block number and offset are easy to compute.
And don't go giving me guff about computers are plenty fast enough to have, say, 1000 disk sectors -- it runs much, much deeper than that.
It's all confusing. There are two billions - the US one (10^9) and the UK one (10^12) which is admittedly not in common use these days.
There's also the kilo/kibi debacle which I think is only useful when the general public isn't involved - I can barely get my father to correlate RAM and disk space (which reminds me - disc (optical, and magneto-optical based on MiniDisc) vs disk (everything else)) let alone the differences between how they're measured. (Son, why does it say I've got a 74GB disk when it says 80GB on the box?)
Not to mention you can go even deeper by investigating the acronyms for bits and bytes - one standard (IEEE 1541) says b=bit and B=byte where another (IEC 60027) says bit=bit and B=byte, because bit is already an abbreviation of Binary Digit. I still refer to 8Mbit/s DSL speeds.
I'm waiting for the Reg to come up with a standard for us all to use for data storage measurement in addition to linguine for distance - such as PoPH (Pictures of Paris Hilton) for small measures, and VMC (Virgin Media Caps) for larger ones
There is no need for MiB or GiB etc. because the B at the end means Bytes, therefore binary, so you don't need to specify that it is a Mega-Binary-Byte, because what the hell else do you make a byte out of? This means that the "and 24" binary rollovers are implied (nay, required) by the numbers that you are counting being binary.
"cheating us out of an ever-growing proportion of what we should be getting"
It's not an ever-increasing proportion. While HD sizes are quoted in GB/GiB, it's consistently ~7% loss. Once we move to TB/TiB (and assuming they use 1000^4 vs 1024^4 bytes), it'll be a consistent ~9%.
The proportion changes each time we up the unit size by an order of magnitude.
I outright refuse to use the travesty of "Kirbybytes" to refer "1024-byte kilobytes" in any kind of use!
As some have noted, the kilo- mega- giga- prefix family has been 1024-based because of technical reasons, and the only idiots that decided to do this otherwise were the HD manufacturers.
Then some zealot purists decided to "end the ambiguity" and declare the Men In Black ... um... I mean the binary prefixes, making all the damn thing worse for actual computer users, and enabling HDD manufacturers to get away with their shoddy practices.
Hey, even real mathematicians/engineers joke about the absurdity:
http://www.xkcd.com/394
and one poster in the forums summed it up all in one line:
"I hate "kibibyte," "gibibyte," etc. It is pedantry of the worst kind."
Pedantry indeed.
I don't care which way you go but adding a 500 GB drive as a 2nd drive in a system should not increase your storage capacity by only 465 GB. The point is not whether or not what HD manufacturers call a GB is right or wrong or whether what the OS calls a GB is right or wrong. The point is, there shouldn't be "two sets of rules". Pick ONE !
Or shalI we start having to buy 20 gallons of gas to fill up an 18 gallon gas tank ?
Those of us blessed with the opportunity to earn an honest living in this industry are more than familiar with consumers/end-users complaining about their 200GB hard disk not showing up as 200GB in Windows. In light of this reality, one's grasp of binary, hard disk "standards", or mathematics in general is inconsequential.
You know what they say: People in glass houses sink sh... sh... sh... ships. - take note
Seagate may have put out more drives, but historically there are a HELL of a lot more Western Digital and Maxtor (pre-acquisition) drives still working after 5 years. Seagate has a horrible reputation for drives that die immediately after the 3-year warranty expires. The average life of a drive is 6 years.
I work at a school with a lot of older computers, with multiple hard drive brands. I can honestly say I have not found a SINGLE working Seagate drive. The Western Digitals and Maxtors are pretty solid (Maxtors especially), but every single Seagate is dead, without fail. I don't even bother to test them anymore. I just throw them out.