[Beowulf] massive parallel processing application required

[Jim Lux]

At 02:37 PM 1/31/2007, Mark Hahn wrote:
>>worked at a e-science project himself. He told me people, especially
>
>is this "e-science" term more popular where you are?  I don't really
>hear it here (Canada, probably NA in general.)  many (most) branches 
>of science are so dependent on computers that it seems redundant and archaic.
>
>>scientists are "very jealous" of their data. And not replying is a 
>>kind way of saying "no". And there's the problem of "who's this guy 
>>wanting my data", "what will he do with it?".
>
>sure - academia is all about publishing, so you don't want someone 
>else to scoop you.


This is an interesting aspect.  All the latest Announcements of 
Opportunity for space research (think cameras taking pictures of 
Mars, or analyzing rocks etc.) have fairly fast time lines(weeks, not 
months or years) for relase of data to general public, and you have 
to put your plans and budgets for public distribution in the 
proposal.  No more holding onto the data, "recalibrating and 
reprocessing",  until everyone gets their dissertations done.

The Dead Sea Scrolls are probably a pretty notorious example of 
"we're still working on it.. when we're done, we'll release 
it"...Amazing what a computer and a concordance can do.

>   but often concerns are much more mundane, like:
>
>         - how long will it transfer this 50GB blob of data?

Think terabytes, for many data sets.  QuikSCAT gives you wind vectors 
for 90% of the world's ice free ocean twice a day on a 25 km 
grid.  Raw radar data coming down from the spacecraft before 
calibration is basically 200 odd measurements/second (each 
measurement consisting of 10 numbers) continuously 24/7.  And, that's 
a low volume sensor.  Something like a SAR doing radar imaging is 
orders of magnitude greater. Sipping from a firehose indeed.


>         - if I provide neurophysiology data from my lab, can I
>         be sued if the subjects' privacy is violated?
>
>         - do I have time to spend explaining how the data is encoded?

That's a biggie... Budgets are always limited, so they tend to do 
just what's needed for the immediate purpose.


>regards, mark hahn.
>_______________________________________________
>Beowulf mailing list, Beowulf at beowulf.org
>To change your subscription (digest mode or unsubscribe) visit 
>http://www.beowulf.org/mailman/listinfo/beowulf

James Lux, P.E.
Spacecraft Radio Frequency Subsystems Group
Flight Communications Systems Section
Jet Propulsion Laboratory, Mail Stop 161-213
4800 Oak Grove Drive
Pasadena CA 91109
tel: (818)354-2075
fax: (818)393-6875