I'm hoping we're going to see more PALFA work units soon. So far I've received only a few. Maybe this is
a limited run to check stability? Does any one know? Bikeman?
The focus is clearly on the GW search (S5R5). The goal for RadioPulsar search is to ultimately spend 20-25% of E@H's total computing power on it. There are some issues still to overcome in reaching this goal, though.
One aspect is that workunit generation works quite differently for ABP1 work than what we are used to. The 'work' is 'produced' in chunks of 628 units, each chunk taking the workunit generator to run for about an hour to pre-process the data, and if something goes wrong it usually gets noticed only at the very end.
Another (short-term) issue is the handling of the first charge of 628 ABP1 workunits that were sent out with a wrong upload URL. Handling (i.e. validating and assimilating) these results takes some manual work until all workunits of this set have been processed. And yes, having sent out only one chunk first looks like the right decision now.
Small number of binary pulsars work units.
)
LOL...
Can't say for sure, but given the cautious nature of the EAH Staff that would be a pretty safe assumption to go on. ;-)
Personally I prefer that to the 'Whip out out into the wild, and see if it blows up' methodology we've seen elsewhere lately. :-D
Alinator
I´ve got not a single one up
)
I´ve got not a single one up to now. Just waiting for the first. Could be that the focus is on all-sky-search?
The focus is clearly on the
)
The focus is clearly on the GW search (S5R5). The goal for RadioPulsar search is to ultimately spend 20-25% of E@H's total computing power on it. There are some issues still to overcome in reaching this goal, though.
One aspect is that workunit generation works quite differently for ABP1 work than what we are used to. The 'work' is 'produced' in chunks of 628 units, each chunk taking the workunit generator to run for about an hour to pre-process the data, and if something goes wrong it usually gets noticed only at the very end.
Another (short-term) issue is the handling of the first charge of 628 ABP1 workunits that were sent out with a wrong upload URL. Handling (i.e. validating and assimilating) these results takes some manual work until all workunits of this set have been processed. And yes, having sent out only one chunk first looks like the right decision now.
BM
BM
Yippiehhhh, i got 2 units.
)
Yippiehhhh, i got 2 units.
http://einstein.phys.uwm.edu/
)
http://einsteinathome.org/workunit/50920983
This one is crunching on one of my nodes right now.
Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.
