Advanced search

Message boards : Number crunching : cross project certificate: not use names like quintillion for numbers

Author Message
root@work
Send message
Joined: 25 Dec 11
Posts: 3
Credit: 17,675,333
RAC: 51
Level
Pro
Scientific publications
watwatwatwatwatwatwatwat
Message 23466 - Posted: 15 Feb 2012 | 10:01:16 UTC


The cross project certificate that can be generated on the user stats web page puts a statement like this:

$USER has contributed 4,697,228 Cobblestones of computation (4.06 quintillion floating-point operations) to the following scientific research projects...

I think that using names like million,...,quintillion etc for numbers is not appropriate for international scientific projects for two reasons:

a) these names represent different numbers in different regions

see http://en.wikipedia.org/wiki/Names_of_large_numbers

e.g. a quintillion in US is 10^18, in Europe it is 10^30 !

b) performance of fast computers (or clusters) is usually expressed in Greek names like Giga, Tera, Peta etc.

see http://en.wikipedia.org/wiki/FLOPS

with that the above statement would read

$USER has contributed 4,697,228 Cobblestones of computation (4.06 exaFLOPS) to the following scientific research projects...

Simba123
Send message
Joined: 5 Dec 11
Posts: 147
Credit: 69,970,684
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 23467 - Posted: 15 Feb 2012 | 10:10:59 UTC - in response to Message 23466.

I do agree with this.

quintillion sounds like a 3 year old making up a name like
gazillion!!

root@work
Send message
Joined: 25 Dec 11
Posts: 3
Credit: 17,675,333
RAC: 51
Level
Pro
Scientific publications
watwatwatwatwatwatwatwat
Message 23468 - Posted: 15 Feb 2012 | 10:28:02 UTC - in response to Message 23466.


$USER has contributed 4,697,228 Cobblestones of computation (4.06 exaFLOPS) to the following scientific research projects...


To avoid different interpretations of FLOPS (sometimes this is read as 'floating point operations per second') we could even use FLOP.

MarkJ
Volunteer moderator
Volunteer tester
Send message
Joined: 24 Dec 08
Posts: 738
Credit: 200,909,904
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 23471 - Posted: 15 Feb 2012 | 20:34:58 UTC

We I suppose if we're being scientific about it we should express it as an exponential like this:

$USER has contributed 4,697,228 cobblestones of computation (4e16 floating point operations)...

Spelling out FLOP is probably better as it explains the unit of measure.
____________
BOINC blog

Toni
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 9 Dec 08
Posts: 1006
Credit: 5,068,599
RAC: 0
Level
Ser
Scientific publications
watwatwatwat
Message 23472 - Posted: 15 Feb 2012 | 20:50:37 UTC - in response to Message 23471.

I agree... it's nonsense to use multiples beyond millions.

JLConawayII
Send message
Joined: 31 May 10
Posts: 48
Credit: 28,893,779
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 23477 - Posted: 16 Feb 2012 | 1:45:42 UTC

Use scientific notation. Problem solved.

Post to thread

Message boards : Number crunching : cross project certificate: not use names like quintillion for numbers

//