# why do computer scientists say 1KB=1024 bytes?!!

• posted

What's so special about 2^10 that computer scientists say it's 1k? The SI system says 1k is 1000. 2^10 is not 1000, it is 1024.

Why not just say 1 KB = 2^9.965784285 bytes (= 1000)? If you're going to nerd out, then nerd out completely, I say.

• posted

It is "close enough" for the marketing assholes too.

My 300 "GB" harddrive manages to have only 270 actual gigs.

• posted

Sure... try saying "the store is 3 kibimeters east of here". Or buying a 5 kibi-watt generator. How many cents per kibiwatt-hour is it in your neck of the woods? :)

And don't even get me started on kb (kilobytes) vs kb (kilobits)! My laptop has a USB 1.1 (not 2.0) port, and I'm trying to calculate how fast I can transfer data. I google it, and discover it can handle 12 mbps! Oh, wait a second, that's megaBITS per second, only 1.5 megaBYTES per second. Geez!

Apparently the 480 mbps USB 2.0 handles only 40 megaBYTES per second. And what's on the package of all those USB 2.0 PCI cards? You guessed it... 480 mbps. Not 40 mbps.

I don't see the hard disk vendors bragging about their 800 gb drives! Oh wait, that's 8 gigaBITs, only 100 gigaBYTEs...

On the other hand, buy some rechargeable NiMH AA batteries, and the package says, "2500 mAh"! Sounds impressive. Two thousand five hundred milli amp hours. I guess 2.5 Ah isn't impressive enough for the marketing folks...

• posted

Correction, USB 2 handles closer to 60 than 40 mBYTES per second. Divided by 12 for some reason instead of by 8. Senior moment.

And according to this, somewhat less than 60 mBytes/s also...

• posted

Well.. the linux kernel have been re-commented and re-documented using kibi and Mibi. So there are people using it practically. And the company I work for is also slowly adopting ki, Mi and Gi since we are in telecommunications and want to avoid coufusion. In telecoms and networking, 10 mega bits per second really does mean 10 million state changes per second, not (approx) 10.485 million. This is mainly because telecoms have never really cared about the binary number system. We're all about signal-to-noise ratios.

You'll want to correct that, kb is always kilobits. kB is kilobytes. Though, sometimes people do forget to press the shift key when they mean bytes.

• posted

Howdy!

Because kb or Kb is KiloBITs - KB is KiloBYTES.

And *ahem* /10 to account for the overhead. And don't forget, a single device can only use a max of 50% of the bandwidth.

It's documented to be actually a bit over 650kilobytes/sec max transfer for USB 1.1 per device.

More like 48KBytes/sec. AND - see above.

Be more than 800gbits for 100gigabytes. Sector info, track info, et al yields more like a 75% to 80% efficiency.

Well - if the originals are 800mAH, then why not 2500mAH for the high powered ones? Same units and all.

RwP

• posted

Curiously enough, my computer measures things that are defined by powers of two. However, my data storage seems to come in things that are defined by powers of ten. I find that odd. Especially because it isn't even consistant - RAM is sold in actual gigabytes and megabytes instead of the marketing version of the gigabyte.

• posted

They say that because a byte is eight bits. It's doesn't really matter what SI, or Physics, IEEE and ASCIII retards have to say about the issue.

• posted

60 millibytes per second?

Some mistake surely, I don't think it takes just over half a year to download a 1MB file via USB 2, but then I never estimated the error bars.

br

• posted

Yes, but a byte has nothing to do with 2 by 4 or elecronics. It;s just that after 100 years of Relativity idiot crap, science is so corrupted by moronic pork barrel politics that the idiots in physics now have to call Capt Kirk on The NASA Rodeo Drive Mel Gibson sub-either tard channel rather than the regular channels.

• posted

You've got it backward... computer folks call 2^10 bytes a KB 2^20 bytes a MB

• posted

(1) When one resorts to profanity and name-calling it is usually a pretty sure indication that the poster has nothing intelligent to say.

(2) How do you know that it "only" has 270 actual Gbytes? Is it because the directory command says that is the amount of free space? Did it occur to you that there is overhead information that must be written to the disk in order to format it? 300 Gbytes, I am sure is the *unformatted* size of the disk. Format information typically takes ~10%.

• posted

rounding

• posted

1024 is close enough to 1000 for jargon. The meaning is determined from the context. If you want to sling around a more exact jargon, then use kibibyte, mebibyte, and gibibyte.
```--
"What are the possibilities of small but movable machines?  They may or
may not be useful, but they surely would be fun to make."```
• posted

K = 1024 k = 1000

is a convention proposed by Tenenbaum. Never caught on though. Computer science have given us another exception:

b = bits B = bytes

Then again, on some very old systems, a byte is not necessarily 8 bits (though modern systems always assumes this). That's why a lot of older documentation talked about 'octets'.

Despite criticisms, kiB, MiB, GiB are catching on. I've started seeing them in tender documents and specifications.

• posted

That's because memory devices have power of 2 addressing (it would actually be harder to make it in powers of 10!). Storage devices have no such limitation, so they ue powers of 10 to make it sound larger. When reported, though, some systems report the power of 2 version (so a

300 GByte drive, to a marketeer - 300 ^E9 - is 279.396 GBytes using power of 2 bases. 1GByte, 2^30).

Cheers

PeteS

• posted

Hey, it isn't just in technical electronics. Ever measured a "2 by 4"?

--Lynn

• posted

First, 1 KB used to be a significant amount of memory. Some systems were designed such that the required amount of memory was estimated at about 1000 bytes. So the hardware guys went to work and designed a memory system. They realized they would need at least 10 discreet address lines so that each of the 1000 memory locations would have a unique address. Then they discovered that it would be cheaper to design the memory system so that the number of memory locations matched the addressing system rather than the other way around. Hence the memory chips had 2^10 locations which very closely matched the 1000 locations that was desired.

• posted

I think I may have seen that one in IP documentation.

An IP(v4) address consists of four octets.

Sure. One hexadecimal character. A convenient unit.

- Randy

• posted

Well, you work for Haliburton, So your opimiom is not only irrelevent, to IEEE, it''s also irrelevent to the Uiniverse

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.