On your account page, middle section, Einstein@home Preferences click there.
Find the profile that you use. In the middle of that profile, it will list all of the different types of work units you will allow to run on your system.
Click edit and modify it so that you have selected Binary Radio Pulsar Survey (Parkes PMPS XT)
Then in the section below that there is another setting that says
Quote:
Run beta/test application versions?
This helps us develop applications, but it may cause jobs to fail on your computer.
Click that also.
Save.
When the next time you download Parkes binary work, it will run the cuda 55 version of the Parkes instead of the 32.
This tends to be faster than the cuda 32 version. It also responds better to multiple work units on each GPU ( at least on the Maxwell cards 750 and 900s, Titans)
But you have to figure out how many you can run per card, That takes time and testing.
I have changed my work to Parkes only and with beta permision. This might be interesting for comparison for total stone earned per day. Its not yet been an entire week and the Free DC bars have not yet stabilized. Looks pretty much like they are adjusting to 120,000 per day.
Are RBPs any more likely to be found in any particular task groups?
Arecibo 19 Oct 2012
Just Because The Space Alien Is Green
Does Not Mean You Should Go
And I do have a small second rig that has a couple of 650s in it but will not power up.
Earlier this year I took it into the shop and it ran fine. Home again and won't power up. Maybe I will use it for a test mule once I kick hard enough.
I'm just tossing this out because I have-had both situations happen to me. ...and you would think that having either thing happen would be rare.
I have had a computer running just fine, no problems at all, that I restarted and it died. After I-can't-tell-you how much computer-work I discovered that the wall socket was bad. I guess it was good enough to keep it running, but not good enough to take the initial power-on surge.
Then, there was a computer I put together from brand new parts. I had to assemble four. One would not come on.
I changed and swapped every component in the machine, cheated the power on switch, etc, and it would not come on. I spent several hours on it. Finally a colleague of mine asks "Did you try a different power cord?" After telling him I thought that was ridiculous (it was a brand new, thick, nice cord straight from the brand new power supply box) I tried a different power cord and machine came on.
If it would run at the shop and won't run at home, it really might be the power cord or the wall socket. As much as that sounds ridiculous, you should check.
Are RBPs any more likely to be found in any particular task groups?
I don't think we have enough information (at least easily interpreted) to know.
Everything I am about to say is probably wrong in some particular:
The Parkes data has been looked-at when it was new and pulsars were found. Einstein at home looked through at least some of it again, in more detail, and found more. IF I understand this (and I probably don't) we are looking at it again looking for binary pulsars with different periods at the same "fine" level that found some pulsars before. With as many pulsars as "should" be out there, it is reasonable to think there are others in the data. Whether our methods will find them is something I know nothing about.
The Gamma-ray pulsar search is, AGAIN if I understand, and I probably don't, a "newish" look. I would think we would be able to find some in there, maybe, if we haven't already exhausted the new data.
The Arecibo? I really don't have any idea. It's new data, but from where? I know this is discoverable, but I don't know exactly how to figure-out if this is a part of the sky where we've already looked or looked for binaries in with an orbital time we're looking-for now.
I quit trying to figure it out, frankly. The information I could find came from so many different threads and was posted at widely differing times that I didn't know if what I was discovering was still true or simultaneously true or all such "old news" that it was now false all the way around.
Honestly, I think it would take a project scientist posting the current goals and describing the current data being crunched as of now for me to feel certain that I know anything.
The best information I've found is in Message 138145. That doesn't tell you the odds of finding a binary pulsar, but it tells you what each data set is about.
VERY good story! Gremlins to right of me. Gremlins to the left. And no Ghost
Busters in the local yellow pages! I am simply happy my four GPU rig powered by a 400 watt Platinum does not crash at 370 watts from the wall plug. As a numismatic guy I keep a Pontius Pilot Lepton (30 AD our calendar struck either Jerusalem or Caesarea) in my wallet.
I am hopeful I will not need a trip to Lourdes.
Arecibo 19 Oct 2012
Just Because The Space Alien Is Green
Does Not Mean You Should Go
I originally signed up to BOINC to look for ET. And SETI has a very fun screen saver. My handle box shows a fine triplet! However, I have been adding actual data into the Drake Equation for decades. Yes. I am that old and its entirely hopeless. Einstein at home has actually found new RBPs, although it is not clear to me how important these individual finds are. Certainly, if we find a seriously powerful one THAT would be by itself an important thing. And even if we simply help populate the sky with additional 'dogs and cats' it can't hurt anything.
And so I am delighted to add what I can add. In my case I have decided to work on efficiency. I have no hope of Giant Stone Numbers! But my original goal was to exceed 35 watts per 10,000 stones. My current rig of Kepler units seems to have done so, but Maxwell has changed the game.
I have ordered a used GTX 750ti from ebay for the single purpose of seeing if it can simply be plugged in to replace my GTX 660 which is a serious energy hog. My original plan was to use four cordless GTX 650s ganged together and fed by a low power I-3. The bottle neck was that the PCIE busses would only support three of them without the fourth one introducing errors. The second problem was simply getting more then two GPUs to work simultaneously. My builder could not do so. [My theory was to keep computer overhead low.]
I managed to get four GTX 650s to gang together, but as stated above, they became unstable. So my plan was to introduce another GTX650 with a cord. None of the ones I tried would gang up. The cordless ones were NVIDIA butI had some EVGA corded versions running in other computers but Windows 7 recognized none of them. I also had an EVGA 660 that DID plug in, and that's what id working now.
Double bubble, toil and trouble. I am looking forward to see if Windows 7 will recognize the used GTX750ti now on order. I am not holding my breath! If it does I might swap out some of the cordless GTC650s. The GTX750 ti actually use less buss power and has many more cuda cores.
Fun and games!
Arecibo 19 Oct 2012
Just Because The Space Alien Is Green
Does Not Mean You Should Go
When you get it, after you plug it in. See if Einstein will use all of the GPUS.
I know others have stated that BOINC will sometimes disable what it considers lesser GPUs in favor of more efficient (newer) ones. In the case of mixing different GPUs like 600s,700s, or 900s
If it does that with your system, you may need to add a cc_config.xml in the BOINC folder to allow Einstein to use all of the GPUs.
Let us know and we can provide one for you. I have not listed it now as I'm not sure if you will need it or not.
... Einstein at home has actually found new RBPs, although it is not clear to me how important these individual finds are.
Then Dr google is your friend. I just did a quick search about pulsar density and by following the internal links came to this little gem about binary systems. I say gem because it's short, concise and in layman's terms, with oodles of links to other things you might be interested in. Note the reference to the Hulse/Taylor binary system and the startling more recent discovery from the Parkes radio telescope data that occurred just a few short years before E@H started sifting through that data as well.
You keep referring to "RBPs". That's OK, we can guess you are referring to the pulsar search using radio emissions rather than other wavelengths. Just be aware that there is also a search (CPU only) using gamma ray emissions. That search uses the LAT telescope on the Fermi satellite and is named 'FGRP4' for Fermi Gamma Ray Pulsar search#4.
The 'radio' search is using data from two different radio telescopes, Arecibo and Parkes. Parkes is important because, being in the southern hemisphere, it can investigate regions that Arecibo can't. The current search using Arecibo data is BRP4 (Binary Radio Pulsar search#4). The current search using Parkes data is BRP6. The Arecibo data is transient in nature because the rate of crunching is faster than the rate of production. Only a fraction of it (BRP4G) is reserved for discrete GPUs. At the moment BRP6 rather than BRP4G is the way to go if you want continuous production.
Whilst the search is called 'Binary Radio Pulsar' most of the discoveries will probably be single pulsars. There are more of them and those that point our way can be found. The most interesting objects are the binary systems - but I'd be very happy to bag any type :-). It always blows my brain when trying to imagine an object about 20km in diameter where a teaspoonful 'weighs' a billion tons which is also rotating, sometimes at incredible speeds. MSPs - millisecond pulsars - the one in the binary system mentioned in the above link is going around on its axis 44 times per second!!
It always blows my brain when trying to imagine an object about 20km in diameter where a teaspoonful 'weighs' a billion tons which is also rotating, sometimes at incredible speeds. MSPs - millisecond pulsars - the one in the binary system mentioned in the above link is going around on its axis 44 times per second!!
I'm sure everyone's mind is blown if they think about it. I conclude gravity is a heavy subject and that's one of the reasons I crunch here.
Thats my kind of discussion! I have been following super massive subjects for a long time. Black holes are interesting because their accretion discs and the law of conservation of angular momentum means almost anything that passes the event horizon is so relativistic that, from our point of view, it never gets much further. Even though the event horizon expands.
I have repeatedly asked science forums whether a hypothetical non rotating neutron star could simply accumulate enough mass to become a black hole. I have never gotten a cogent answer. And of course there are no such hypotheticals. But I digress.
And I agree that gravity is the heavy subject!
Arecibo 19 Oct 2012
Just Because The Space Alien Is Green
Does Not Mean You Should Go
On your account page, middle
)
On your account page, middle section, Einstein@home Preferences click there.
Find the profile that you use. In the middle of that profile, it will list all of the different types of work units you will allow to run on your system.
Click edit and modify it so that you have selected Binary Radio Pulsar Survey (Parkes PMPS XT)
Then in the section below that there is another setting that says
Click that also.
Save.
When the next time you download Parkes binary work, it will run the cuda 55 version of the Parkes instead of the 32.
This tends to be faster than the cuda 32 version. It also responds better to multiple work units on each GPU ( at least on the Maxwell cards 750 and 900s, Titans)
But you have to figure out how many you can run per card, That takes time and testing.
Zalster
I have changed my work to
)
I have changed my work to Parkes only and with beta permision. This might be interesting for comparison for total stone earned per day. Its not yet been an entire week and the Free DC bars have not yet stabilized. Looks pretty much like they are adjusting to 120,000 per day.
Are RBPs any more likely to be found in any particular task groups?
Arecibo 19 Oct 2012
Just Because The Space Alien Is Green
Does Not Mean You Should Go
RE: And I do have a small
)
I'm just tossing this out because I have-had both situations happen to me. ...and you would think that having either thing happen would be rare.
I have had a computer running just fine, no problems at all, that I restarted and it died. After I-can't-tell-you how much computer-work I discovered that the wall socket was bad. I guess it was good enough to keep it running, but not good enough to take the initial power-on surge.
Then, there was a computer I put together from brand new parts. I had to assemble four. One would not come on.
I changed and swapped every component in the machine, cheated the power on switch, etc, and it would not come on. I spent several hours on it. Finally a colleague of mine asks "Did you try a different power cord?" After telling him I thought that was ridiculous (it was a brand new, thick, nice cord straight from the brand new power supply box) I tried a different power cord and machine came on.
If it would run at the shop and won't run at home, it really might be the power cord or the wall socket. As much as that sounds ridiculous, you should check.
RE: Are RBPs any more
)
I don't think we have enough information (at least easily interpreted) to know.
Everything I am about to say is probably wrong in some particular:
The Parkes data has been looked-at when it was new and pulsars were found. Einstein at home looked through at least some of it again, in more detail, and found more. IF I understand this (and I probably don't) we are looking at it again looking for binary pulsars with different periods at the same "fine" level that found some pulsars before. With as many pulsars as "should" be out there, it is reasonable to think there are others in the data. Whether our methods will find them is something I know nothing about.
The Gamma-ray pulsar search is, AGAIN if I understand, and I probably don't, a "newish" look. I would think we would be able to find some in there, maybe, if we haven't already exhausted the new data.
The Arecibo? I really don't have any idea. It's new data, but from where? I know this is discoverable, but I don't know exactly how to figure-out if this is a part of the sky where we've already looked or looked for binaries in with an orbital time we're looking-for now.
I quit trying to figure it out, frankly. The information I could find came from so many different threads and was posted at widely differing times that I didn't know if what I was discovering was still true or simultaneously true or all such "old news" that it was now false all the way around.
Honestly, I think it would take a project scientist posting the current goals and describing the current data being crunched as of now for me to feel certain that I know anything.
The best information I've found is in Message 138145. That doesn't tell you the odds of finding a binary pulsar, but it tells you what each data set is about.
VERY good story! Gremlins to
)
VERY good story! Gremlins to right of me. Gremlins to the left. And no Ghost
Busters in the local yellow pages! I am simply happy my four GPU rig powered by a 400 watt Platinum does not crash at 370 watts from the wall plug. As a numismatic guy I keep a Pontius Pilot Lepton (30 AD our calendar struck either Jerusalem or Caesarea) in my wallet.
I am hopeful I will not need a trip to Lourdes.
Arecibo 19 Oct 2012
Just Because The Space Alien Is Green
Does Not Mean You Should Go
Very good discussion. I
)
Very good discussion.
I originally signed up to BOINC to look for ET. And SETI has a very fun screen saver. My handle box shows a fine triplet! However, I have been adding actual data into the Drake Equation for decades. Yes. I am that old and its entirely hopeless. Einstein at home has actually found new RBPs, although it is not clear to me how important these individual finds are. Certainly, if we find a seriously powerful one THAT would be by itself an important thing. And even if we simply help populate the sky with additional 'dogs and cats' it can't hurt anything.
And so I am delighted to add what I can add. In my case I have decided to work on efficiency. I have no hope of Giant Stone Numbers! But my original goal was to exceed 35 watts per 10,000 stones. My current rig of Kepler units seems to have done so, but Maxwell has changed the game.
I have ordered a used GTX 750ti from ebay for the single purpose of seeing if it can simply be plugged in to replace my GTX 660 which is a serious energy hog. My original plan was to use four cordless GTX 650s ganged together and fed by a low power I-3. The bottle neck was that the PCIE busses would only support three of them without the fourth one introducing errors. The second problem was simply getting more then two GPUs to work simultaneously. My builder could not do so. [My theory was to keep computer overhead low.]
I managed to get four GTX 650s to gang together, but as stated above, they became unstable. So my plan was to introduce another GTX650 with a cord. None of the ones I tried would gang up. The cordless ones were NVIDIA butI had some EVGA corded versions running in other computers but Windows 7 recognized none of them. I also had an EVGA 660 that DID plug in, and that's what id working now.
Double bubble, toil and trouble. I am looking forward to see if Windows 7 will recognize the used GTX750ti now on order. I am not holding my breath! If it does I might swap out some of the cordless GTC650s. The GTX750 ti actually use less buss power and has many more cuda cores.
Fun and games!
Arecibo 19 Oct 2012
Just Because The Space Alien Is Green
Does Not Mean You Should Go
When you get it, after you
)
When you get it, after you plug it in. See if Einstein will use all of the GPUS.
I know others have stated that BOINC will sometimes disable what it considers lesser GPUs in favor of more efficient (newer) ones. In the case of mixing different GPUs like 600s,700s, or 900s
If it does that with your system, you may need to add a cc_config.xml in the BOINC folder to allow Einstein to use all of the GPUs.
Let us know and we can provide one for you. I have not listed it now as I'm not sure if you will need it or not.
Zalster
RE: ... Einstein at home
)
Then Dr google is your friend. I just did a quick search about pulsar density and by following the internal links came to this little gem about binary systems. I say gem because it's short, concise and in layman's terms, with oodles of links to other things you might be interested in. Note the reference to the Hulse/Taylor binary system and the startling more recent discovery from the Parkes radio telescope data that occurred just a few short years before E@H started sifting through that data as well.
You keep referring to "RBPs". That's OK, we can guess you are referring to the pulsar search using radio emissions rather than other wavelengths. Just be aware that there is also a search (CPU only) using gamma ray emissions. That search uses the LAT telescope on the Fermi satellite and is named 'FGRP4' for Fermi Gamma Ray Pulsar search#4.
The 'radio' search is using data from two different radio telescopes, Arecibo and Parkes. Parkes is important because, being in the southern hemisphere, it can investigate regions that Arecibo can't. The current search using Arecibo data is BRP4 (Binary Radio Pulsar search#4). The current search using Parkes data is BRP6. The Arecibo data is transient in nature because the rate of crunching is faster than the rate of production. Only a fraction of it (BRP4G) is reserved for discrete GPUs. At the moment BRP6 rather than BRP4G is the way to go if you want continuous production.
Whilst the search is called 'Binary Radio Pulsar' most of the discoveries will probably be single pulsars. There are more of them and those that point our way can be found. The most interesting objects are the binary systems - but I'd be very happy to bag any type :-). It always blows my brain when trying to imagine an object about 20km in diameter where a teaspoonful 'weighs' a billion tons which is also rotating, sometimes at incredible speeds. MSPs - millisecond pulsars - the one in the binary system mentioned in the above link is going around on its axis 44 times per second!!
Cheers,
Gary.
RE: It always blows my
)
I'm sure everyone's mind is blown if they think about it. I conclude gravity is a heavy subject and that's one of the reasons I crunch here.
Hi Gary! Thats my kind of
)
Hi Gary!
Thats my kind of discussion! I have been following super massive subjects for a long time. Black holes are interesting because their accretion discs and the law of conservation of angular momentum means almost anything that passes the event horizon is so relativistic that, from our point of view, it never gets much further. Even though the event horizon expands.
I have repeatedly asked science forums whether a hypothetical non rotating neutron star could simply accumulate enough mass to become a black hole. I have never gotten a cogent answer. And of course there are no such hypotheticals. But I digress.
And I agree that gravity is the heavy subject!
Arecibo 19 Oct 2012
Just Because The Space Alien Is Green
Does Not Mean You Should Go