13 Aug
2004
13 Aug
'04
7:27 a.m.
Am 13 Aug 2004 14:55 meinte John Lawson:
On Thu, 12 Aug 2004, Stan Barr wrote:
Hi,
ben franchuk <bfranchuk(a)jetnet.ab.ca> said:
> Stan Barr wrote:
>> Work is still in progress. Shizuka University in Japan recently demoed
[snippage]
>> What is too slow. Paper Tape is slow. :)
> They don't say how fast, but imagine how long it would take to read
> 2000Gb at paper tape speeds :-) No need to imagine - Calculate:
Lets say that the Paper Tape is in ASCII format,
then each character is
a Byte - and that the Reader is a fairly advanced optical device running
at a conservative 1200 bytes per minute.
2000 GB (2 TB) is 2*10^12
so (2*10^12)/1200 = 1.66666667*10^9 minutes to read 2
TB
/60 = 27,777,777.8 hours
/24 = 1,157,407.41 days
/365 = 3171 years
presupposing a Really Good reader that could operate that long w/out any
maintenance or downtime.... probably Sellam has a couple of these on his
Shelves.
NOW: 10 KB of 5-mil thick punched paper tape on a
3/4-inch hub with a
punch width of .060 and an inter-character width of .060 makes a diameter
of.... oops my brain just exploded - sorry.
But damn! that would be an impressive wheel of
paper!!!!
Nowwaitaminute.... Impressive Paper roll? that reminds me
of news paper printing, and AFAIR they use speeds of above
40 km/h (25 mph) at at the press in one newspaper here in Munich
(Sueddeutsche Zeitung)... With way less quality paper than
paper tape used to be.
Now .. let's for now assume a bit hole is about 1 mm in diameter,
and the gap between is of equal size, thus one byte every 2mm.
now, assumeing a speed of 40 km/h this gives 40,000,000/2 Bytes/h
or 20 MByte/h transfer rate (or with around 333 KByte/min or 5 k/s
almost 300 times the speed you assumed). The problem is still the
paper transport, not optical sensing, since even an 8 Bit controler
is be able to shovel awy the data.
2*10^12/2*10^7
... gee that's easy to cancel down ...
1*10^5 hours
or just about 14 Month
Hey, that's fast!
Average access time of just 7 Month - large enough paper roll assumed.
(2*10^12*2^-3m = 4 mil km, or 11 times the distance eath/moon - hey,
waitaminute, this means our reader can eat it's way to the moon in
just a month :)
Gruss
H.
It's your fault, John, you got me going :)
--
VCF Europa 6.0 am 30.April und 01.Mai 2005 in Muenchen
http://www.vcfe.org/