Smart Report in Todays Irish Times , 4 Dublin Exchanges 'Live'

carbsy · 27-03-2005 6:04pm

WizZard wrote:

I meant the 512Mb parts

Oh yeah, wasn't looking over there.

CyberGhost · 27-03-2005 6:59pm

Hedgehoggski wrote:

Yeeks! :eek:

Ok, it seems there are those who could use my 2c

First of all, I think the following use of letters is a "bit" of a convention :
Mb=megabit
MB=megabyte
Kb=kilobit
KB=kilobyte

also:
8 bits = 1 byte
1024 bytes = 1 kilobyte
1024 kilobytes = 1 megabyte

(in bit talk, kilo means 1024 rather than 1000 since 1024 is a power of 2 and 1000 is not. )

now from the above
512 kilobits = 512 x 1024 bits = 524288 bits = 524288/8 bytes = (524288/8)/1024 kilobytes = 64 kilobytes

similarly
1024 kilobits = 1024 x 1024 bits = 1048576 bits = 1048576/8 bytes = (1048576/8)/1024 kilobytes = 128 kilobytes

and
2048 kilobits = 2048 x 1024 bits = 2097152 bits = 2097152 / 8 bytes = (2097152 / 8)/1024 kilobytes = 256 kilobytes

in other words,
your 56Kb connection can download a theoretical max of 7KB per second
your 512Mb connection can download a theoretical max of 64KB per second
your 1024Mb connection can download a theoretical max of 128KB per second
your 2048Mb connection can download a theoretical max of 256KB per second

Hope the above helps clarify matters a little.

HH

nice explanation! Thanks

but I still think there is no way we will download at 256KB/s

215-220, being max

I hope I'm wrong again though

Stephen · 27-03-2005 9:16pm

That's correct, as you have to take the PPPoE/ATM overheads into account. You'll never actually download at the max theoretical line speed.

OFDM · 28-03-2005 12:48pm

2Mbps = 2,000,000 bps
8 bits = 1 byte

2,000,000 bps / 8 = 250,000 bytes per second (B/s)
1 kB = 1024 bytes

=> 250,000 B/s / 1024 = 244.14 kB/s

That's the theoretical download limit BEFORE headers, so estimates of 215-220 in previous posts are correct.

Hedgehoggski · 22-03-2010 8:29am

Yeah, oops.. meant Kb instead of Mb on the left column
stupid late night finger trouble! Thanks for the correction and for being very welcoming and nice about what was a rather careless error folks!

Sorry for delay in getting back(!). Been busy!

Of course as others rightly pointed out, all these figures are before any necessary overhead such as error correction bits, and other metadata in the frames which in practice would reduce the actual max user data transfer possible. It is difficult to quantify the size of the average "bite" these housekeeping bits would take out of your data stream as amongst other things, it is dependent on your individual settings, but 215-221KB for every 256KB sent is probably as good a guess as any.

I'm not certain telecoms companies use K=1000. It's likely they do though as it's in their favour. It sneakily Puffs up their data rates a bit by mixing up the use of the conventions of K=1024 (for good reasons!) in the computer world and K=1000 in the "selling services and hardware to computer users" world

Naturally if they do then revise post accordingly by replacing each "1024" by "1000" . The proper user calculation should produce a result in computer KB though, i.e. K=1024, because every KB of the files you send and recieve from a computer uses K=1024 as it's convention

Unfortunately Hard drive companies certainly do use K=1000 though so beware!