Few Questions
25 Dec 2005 12:26:14 UTC
Topic 190445
(moderation:
I'm just curious...
1. is E@H planned to be active project only until first gravitational wave is detected or is it planned that E@H stay active after first detection searching for other pulsar GW sources?
2. Is there any plan to extend E@H search to other sources of GWs beside pulsars?
3. Once Advanced LIGO become reality, is it planned that E@H crunch for Advanced LIGO as well?
4. Is there any chance that some of LISA's data will be crunched by E@H? (I know it's a different project and 5-6 years until it's start, but still...)
Thanks.
Few Questions
)
The detection of a pulsar would elevate E@H from a search to a survey. So while the detailed analysis of such a source would not require the raw crunching power of E@H defining such a source's proper place in the cosmos would.
Thanks Mark for your
)
Thanks Mark for your comment.
It would be very interesting if we could get some input from Project staff here. Thanks.
Edo
RE: I'm just
)
I'd say it's yes to all. I can't imagine the E@H horsepower not being integral to such future attempts. However, at the moment I expect the LIGO collaboration is quite focussed upon achieving the first ( verifiable ) gravitational wave detection ever! ( The Hulse/Taylor pulsar spin down has shown the existence of such radiation at source ).
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Edo, We (meaning the whole
)
Edo,
We (meaning the whole LIGO Science Collaboration, not just those involved with E@H) have been focused on trying to get to the first detection and haven't thought nearly as much about what comes afterward. But it's hard to imagine taking E@H offline after we've found something. We'll always want to analyze more data to search for other things. And analyzing more data on that first pulsar will help too - for example, analyzing a pulsar for a couple of years can tell you how far away it is. (For shorter stretches of data that information is tangled up with how high the hill is.)
There has been discussion from time to time in the Collaboration of using E@H for sources other than pulsars, but it seems unlikely. The basic driving fact is that the all-sky pulsar search is the most computationally demanding one out there, so it's the natural candidate to use the big power you're all donating to E@H.
By the time advanced LIGO comes around ... it's hard to say what computing will look like 5 or more years from now, but it's hard to believe something like E@H won't still be around. And if it is, we'll still need it, mainly for the pulsar searches.
LISA is another ballgame entirely. First, it'll probably be delayed beyond the advanced LIGO time frame. Second, it'll be run by NASA, and they have a whole different set of rules and expectations which I'm not very familiar with. From what I do know, they tend to do things in-house as much as possible. But some of their data analysis needs will be pretty intense, and whatever descendant of BOINC is around 10 years from now would be very useful I'm sure.
Hope this helps,
Ben
Ben, Thank you very much
)
Ben,
Thank you very much for your comments. Everything sounds very interesting and exciting.
Edo
RE: ... The basic driving
)
Nahhh, don't believe that.
The computational appetite of scientists grows according to the computing power available. The current search may have been just beyond reach when E@h was first started, may still be now, but by the time E@h becomes obsolete gravy wave astronomers will have found some even more computationally demanding app.
And the most likely thing to stimulate those new questions will be a handful of solid pulsar observations from E@h.
Like when Bell discovered the pulsar nobdy thought of using them for anything else, the big puzzle was to work out what they were at all (and they were SETI candidates for a while!) - now pulsars as a tool for calibrating universal expansion. Who knows what the theorticians will want to do with the measuremnts when we get beyond the point where "is it a real one" is the issue.
Just wait - once there are about six known GW sources someone will notice that they differ in some ways and are similar in others. Someone will think they know why, and someone else will disagree. That is when round 1 starts, all of this is prelude...
And if E@h closes, there will be other GW projects coming onstream by then to take up the slack, just like Rosetta (a biochem prohect) got the legacy of the Find-a-drug crunchers when FaD closed recently.
River~~
~~gravywavy
Very interesting
)
Very interesting observations, River.
Edo
gravywavy
)
gravywavy wrote:
The thing about the all-sky pulsar search is that it can take as much computing power as you can throw at it. The key point is that pulsars are always "on", which means you want to integrate the longest stretches of data you can. This is in contrast to other sources like supernovae, which go "blip" and then are quiet. And the computational cost rises very steeply as you integrate longer stretches of data, so your available CPU power is what limits you. Searches for short-lived signals will never have the same computational cost as long searches for unknown continuous signals.
The length of the data stretches analyzed is limited by the computational power available rather than by how much data is available. S3 and S4 were a couple months each, but the "einstein" workunits were only looking at 10-hour stretches of data and "albert" workunits are looking at 30-hour stretches of data. The fact that we can afford that with albert is because more people have donated CPU time (and we played some better math tricks with the analysis), and we're still looking at this search taking several times longer than the older ones did.
I suspect that things will indeed get very lively when we detect a pulsar. And when we've got a catalogue of them, we can start doing population studies. But the hard part is always finding the things in the first place.
Ben