I noticed something interesting when I detached my machines, which is key to this problem, but largly ignored, in favour of the "slow machines" argument.
When I detached E@H, boinc proceeded to download work in other projects to fill the gap, loading up on enough work to meet the "connect every x days" criteria in the remaining projects.
Then when I attached a new project on this machine to replace E@H, boinc proceeded to download exactly x days of work in that project. (Im using x becase I dont want you people to get hung up on the number of days - thats not the point).
On my laptop for example (usually runs 2 projects) I now have is 2x days worth of work to process, or double what I expected. Boinc (version 4.25) seems to download enough work to fill x days for each project, yet when you consider that on this machine the work is shared 50/50 on a single processor, it is clear that the total work will take twice the time to process.
For example, if you have the setting at 2 days and run 4 projects, you end up with a total of 8 days of work in the queue for a single processor.
People have commented that I seem to have too much work in the queue. This is why, and it certainly doesnt help anyone meet a tight deadline. With other projects, there is more slack in the deadline and this does not become a problem.
I guess this behavoir makes sense since if you ran CPDN, the machine would never download work for any other project if the x days was taken across all projects as I expected, but it is something that is not clear or immediately obvious.
Ive said I was leaving and dont mean to keep posting here, but I thought this was relevent.
I wonder why people are so stubborn with setting of x days (where x>1)? I have set to connect every 0.25 days and never miss a deadline and never get out of work either. And in fact, running projects connect maybe 3 times a day to report and get new work, while projects that are down contact schedulers according to retry rules, that, they seem, are independent of connect every x.y days setting. (Running CPDN, S@H, PP@H, LHC@H, LHC@H-alpha, BURP, crashcollection and resourcemeasurment - the later 3 are down ATM)
> Since all graphics and the screensaver are using OpenGL, they are taking
> actual cycles away from crunching. It doesn't matter if you use a slow PC or
> fast PC for that. Yet if you don't use the screensaver, don't allow
> Windows/your other OS to use a screensaver and just manually shut down your
> monitor when you leave your desk, you will notice an increase in crunching
> speed.
Not true... The whole idea of switching to OpenGL was to bump the graphics work on to the Graphics Card as opposed to being done on the CPU. As the crunching doesn't use GPU cycles YET (There has been talk as to how practical this could be.) it, should make minimal difference.
>With SETI@Home the order in which work is completed is unimportant. Not
>knowing the science behind Einstein@Home I cannot say for sure that they have
>the same luxury. LHC@Home for sure does not. Projects such as these may need
>the work returned in order, and within a time constraint that does not allow
>for long delays between issuing the work and the reception of the answer.
If LHC have the same urgent work requirement issues that you imply are affecting the Einstein project, how come they can afford the time to give me a fortnight to do the work?? Doesn't add up in my book!! If LHC can give us crunchers an achievable deadline, why can't Einstein??
>The computer I am running Einstein@Home on is a good computer. I've spent
>hundreds of dollars upgrading it, and the only part that is slower than a new
>computer is the processor. The processor is 448 MHz, which is good for a
>Pentium II. It isn't like I pulled an ancient computer out of the garbage dump
>and started running Einstein@Home on it. My computer easily runs Windows XP
>and almost all of the newest applications and games. Running Einstein@Home is
>an easy task for the computer, I still have plenty of RAM and hard disk space
>left. I just think the deadline should be increased so the processor has time
>to process all of the data.
You are saying that this is a games machine?? Try running something like Half Life 2 at full detail in something like 1280x1024! I'm sorry to say Josh that a 450MHz PII wouldn't even grace my room as a doorstop leave alone a functional machine. For £350 - £400 I could pick up a brand spanking new P4 or Semperon machine. For another £70 odd, I can pick up 128MB 8xAGP GFX card. Your specs are ANCIENT!!! Face it!! For a crunching machine however, I'd put it to work in a "Junker Farm" until it died.
[EDIT] For those of you who are interested, the 9th April WU I was talking about, did make it in time thanks to some individual project stalling!! [EDIT]
To add to what I was saying... The 1900+ in question has LHC WUs, the earliest of which is due on the 24th... Exactly 14 days time... A fortnight... Why can't Einstein afford to do this if LHC can if the work being done is equaly time critical?
> To add to what I was saying... The 1900+ in question has LHC WUs, the earliest
> of which is due on the 24th... Exactly 14 days time... A fortnight... Why
> can't Einstein afford to do this if LHC can if the work being done is equaly
> time critical?
I think Todd is right. It is time to quit.
Neither he, nor you, seem to want to understand what many of us are trying to explain. Because LHC@Home has a 14 day deadline in no way does it imply that Einstein@Home's requirements can be met with the same deadline.
Todd said he does not want to dictate, yet, there is his never ending insistance that if he were accomdated that world would be safe for democracy. The fact that there are people out there that want to contribute but cannot is a fact of life. Not all participants can join all projects. The fact that they would add to the resources available is true, but hardly vital. The fact of the matter is that right now we almost have more participants than needed.
LHC@Home came to the table with what they thought would be 12-18 months of work. We ran them dry in 4, and that with a very restricted participant base. Now they run an expiriment, analyze the data and come back in a month or two with a new expiriment. The fact that more people could join is nice, but not really necessary. When SETI@Home Classic shuts down it is possible that the project is going to have a hard time producing work to keep everyone running.
Like gravity, the 7 day deadline for some projects just is a fact of life, for other projects it is 14 days, and there are people who have complained about that being too short. BOINC's scheduling and estimating are not as good as we would like, and there is work being done to highlight and solve these issues. In the mean time, we have suggested ways you and others can participate, and be successful ...
> Todd said he does not want to dictate, yet, there is his never ending
> insistance that if he were accomdated that world would be safe for democracy.
I am beginning to tire of being mocked in this forum, especially after I have already stated that I intend to leave, and in response to somebody elses post.
Paul you and I already know we disagree on various subjects. Please lets leave it at that and not make things worse.
A Pentium II may not sound like that great of a processor, but the computer runs really well. I don't need to be told what using a fast computer is like because I recently bought a new Compaq computer that could probably race by what most of you are calling contemporary machines. The point isn't how long it takes for the Pentium II to process something, the point is that it is capable of processing it, and helping out the Einstein@Home project.
It is funny that someone said the Pentium II computer wouldn't be able to run anything above Need For Speed 2, because I have Need For Speed 3, and it works great even when the graphics are set to best quality. A lot of my games have higher requirements than Need For Speed 3, and they all work fine. SimCity 3000 was a little slow, but I installed some more RAM, and now that works great. Some of my games won't install on the computer because it says my computer is too slow, but when I manually copy the files from the CD, they all run at a good speed. Counter-Strike and Half Life run smoothly on my Pentium II.
Anyway, the purpose of starting this thread wasn't to talk about the games my system can run. You all should think about what would happen if every computer in the world that is capable of running Einstein@Home, but can't make the deadlines started running Einstein@Home. I'm sure the number of results returned each day would multiply by the hundreds. SETI@HOME was started back when a Pentium II was known as a fast machine. It may have taken a while for them to return the results, but thousands of them working together processed so much that SETI@HOME's server kept overloading. The number of results that were returned was amazing. Now, these same machines are being turned away from grid projects by deadlines that have no point. If the computer finds a gravitational wave in a work unit, why does it matter if it doesn't return the result in two weeks instead of one? Does Einstein@Home plan on flying across the galaxy to analyze the wave before it moves away? I don't think so.
> I am beginning to tire of being mocked in this forum, especially after I have
> already stated that I intend to leave, and in response to somebody elses
> post.
>
> Paul you and I already know we disagree on various subjects. Please lets leave
> it at that and not make things worse.
I hear ya, I have an AMD K6-III 400 that works perfectly fine as a web browser and word processor. It's a great little machine within the limits of what it is used it for. But, since i have other machines, it doesn't do any crunching, because to me, it's not worth spending money on the electricity to have it crunch, it just takes too long. Now if it was my only machine, that would be a different story. I'd prolly set it up to run just one project and let it go 24x7.
Obsolescence is inevitable though, the line between what's obsolete and what's viable just keeps moving forward as new stuff comes out and the line has to get drawn somewhere I guess.
> A Pentium II may not sound like that great of a processor, but the computer
> runs really well. I don't need to be told what using a fast computer is like
> because I recently bought a new Compaq computer that could probably race by
> what most of you are calling contemporary machines. The point isn't how long
> it takes for the Pentium II to process something, the point is that it is
> capable of processing it, and helping out the Einstein@Home project.
I noticed something
)
I noticed something interesting when I detached my machines, which is key to this problem, but largly ignored, in favour of the "slow machines" argument.
When I detached E@H, boinc proceeded to download work in other projects to fill the gap, loading up on enough work to meet the "connect every x days" criteria in the remaining projects.
Then when I attached a new project on this machine to replace E@H, boinc proceeded to download exactly x days of work in that project. (Im using x becase I dont want you people to get hung up on the number of days - thats not the point).
On my laptop for example (usually runs 2 projects) I now have is 2x days worth of work to process, or double what I expected. Boinc (version 4.25) seems to download enough work to fill x days for each project, yet when you consider that on this machine the work is shared 50/50 on a single processor, it is clear that the total work will take twice the time to process.
For example, if you have the setting at 2 days and run 4 projects, you end up with a total of 8 days of work in the queue for a single processor.
People have commented that I seem to have too much work in the queue. This is why, and it certainly doesnt help anyone meet a tight deadline. With other projects, there is more slack in the deadline and this does not become a problem.
I guess this behavoir makes sense since if you ran CPDN, the machine would never download work for any other project if the x days was taken across all projects as I expected, but it is something that is not clear or immediately obvious.
Ive said I was leaving and dont mean to keep posting here, but I thought this was relevent.
I wonder why people are so
)
I wonder why people are so stubborn with setting of x days (where x>1)? I have set to connect every 0.25 days and never miss a deadline and never get out of work either. And in fact, running projects connect maybe 3 times a day to report and get new work, while projects that are down contact schedulers according to retry rules, that, they seem, are independent of connect every x.y days setting. (Running CPDN, S@H, PP@H, LHC@H, LHC@H-alpha, BURP, crashcollection and resourcemeasurment - the later 3 are down ATM)
[edit] Computer is a P4 1.8GHz [/edit]
Happy crunching,
> Since all graphics and the
)
> Since all graphics and the screensaver are using OpenGL, they are taking
> actual cycles away from crunching. It doesn't matter if you use a slow PC or
> fast PC for that. Yet if you don't use the screensaver, don't allow
> Windows/your other OS to use a screensaver and just manually shut down your
> monitor when you leave your desk, you will notice an increase in crunching
> speed.
Not true... The whole idea of switching to OpenGL was to bump the graphics work on to the Graphics Card as opposed to being done on the CPU. As the crunching doesn't use GPU cycles YET (There has been talk as to how practical this could be.) it, should make minimal difference.
>With SETI@Home the order in which work is completed is unimportant. Not
>knowing the science behind Einstein@Home I cannot say for sure that they have
>the same luxury. LHC@Home for sure does not. Projects such as these may need
>the work returned in order, and within a time constraint that does not allow
>for long delays between issuing the work and the reception of the answer.
If LHC have the same urgent work requirement issues that you imply are affecting the Einstein project, how come they can afford the time to give me a fortnight to do the work?? Doesn't add up in my book!! If LHC can give us crunchers an achievable deadline, why can't Einstein??
>The computer I am running Einstein@Home on is a good computer. I've spent
>hundreds of dollars upgrading it, and the only part that is slower than a new
>computer is the processor. The processor is 448 MHz, which is good for a
>Pentium II. It isn't like I pulled an ancient computer out of the garbage dump
>and started running Einstein@Home on it. My computer easily runs Windows XP
>and almost all of the newest applications and games. Running Einstein@Home is
>an easy task for the computer, I still have plenty of RAM and hard disk space
>left. I just think the deadline should be increased so the processor has time
>to process all of the data.
You are saying that this is a games machine?? Try running something like Half Life 2 at full detail in something like 1280x1024! I'm sorry to say Josh that a 450MHz PII wouldn't even grace my room as a doorstop leave alone a functional machine. For £350 - £400 I could pick up a brand spanking new P4 or Semperon machine. For another £70 odd, I can pick up 128MB 8xAGP GFX card. Your specs are ANCIENT!!! Face it!! For a crunching machine however, I'd put it to work in a "Junker Farm" until it died.
[EDIT] For those of you who are interested, the 9th April WU I was talking about, did make it in time thanks to some individual project stalling!! [EDIT]
To add to what I was
)
To add to what I was saying... The 1900+ in question has LHC WUs, the earliest of which is due on the 24th... Exactly 14 days time... A fortnight... Why can't Einstein afford to do this if LHC can if the work being done is equaly time critical?
> To add to what I was
)
> To add to what I was saying... The 1900+ in question has LHC WUs, the earliest
> of which is due on the 24th... Exactly 14 days time... A fortnight... Why
> can't Einstein afford to do this if LHC can if the work being done is equaly
> time critical?
I think Todd is right. It is time to quit.
Neither he, nor you, seem to want to understand what many of us are trying to explain. Because LHC@Home has a 14 day deadline in no way does it imply that Einstein@Home's requirements can be met with the same deadline.
Todd said he does not want to dictate, yet, there is his never ending insistance that if he were accomdated that world would be safe for democracy. The fact that there are people out there that want to contribute but cannot is a fact of life. Not all participants can join all projects. The fact that they would add to the resources available is true, but hardly vital. The fact of the matter is that right now we almost have more participants than needed.
LHC@Home came to the table with what they thought would be 12-18 months of work. We ran them dry in 4, and that with a very restricted participant base. Now they run an expiriment, analyze the data and come back in a month or two with a new expiriment. The fact that more people could join is nice, but not really necessary. When SETI@Home Classic shuts down it is possible that the project is going to have a hard time producing work to keep everyone running.
Like gravity, the 7 day deadline for some projects just is a fact of life, for other projects it is 14 days, and there are people who have complained about that being too short. BOINC's scheduling and estimating are not as good as we would like, and there is work being done to highlight and solve these issues. In the mean time, we have suggested ways you and others can participate, and be successful ...
> Todd said he does not want
)
> Todd said he does not want to dictate, yet, there is his never ending
> insistance that if he were accomdated that world would be safe for democracy.
I am beginning to tire of being mocked in this forum, especially after I have already stated that I intend to leave, and in response to somebody elses post.
Paul you and I already know we disagree on various subjects. Please lets leave it at that and not make things worse.
A Pentium II may not sound
)
A Pentium II may not sound like that great of a processor, but the computer runs really well. I don't need to be told what using a fast computer is like because I recently bought a new Compaq computer that could probably race by what most of you are calling contemporary machines. The point isn't how long it takes for the Pentium II to process something, the point is that it is capable of processing it, and helping out the Einstein@Home project.
It is funny that someone said the Pentium II computer wouldn't be able to run anything above Need For Speed 2, because I have Need For Speed 3, and it works great even when the graphics are set to best quality. A lot of my games have higher requirements than Need For Speed 3, and they all work fine. SimCity 3000 was a little slow, but I installed some more RAM, and now that works great. Some of my games won't install on the computer because it says my computer is too slow, but when I manually copy the files from the CD, they all run at a good speed. Counter-Strike and Half Life run smoothly on my Pentium II.
Anyway, the purpose of starting this thread wasn't to talk about the games my system can run. You all should think about what would happen if every computer in the world that is capable of running Einstein@Home, but can't make the deadlines started running Einstein@Home. I'm sure the number of results returned each day would multiply by the hundreds. SETI@HOME was started back when a Pentium II was known as a fast machine. It may have taken a while for them to return the results, but thousands of them working together processed so much that SETI@HOME's server kept overloading. The number of results that were returned was amazing. Now, these same machines are being turned away from grid projects by deadlines that have no point. If the computer finds a gravitational wave in a work unit, why does it matter if it doesn't return the result in two weeks instead of one? Does Einstein@Home plan on flying across the galaxy to analyze the wave before it moves away? I don't think so.
> I am beginning to tire of
)
> I am beginning to tire of being mocked in this forum, especially after I have
> already stated that I intend to leave, and in response to somebody elses
> post.
>
> Paul you and I already know we disagree on various subjects. Please lets leave
> it at that and not make things worse.
Um, where did I mock you?
> Um, where did I mock
)
> Um, where did I mock you?
I dont believe I actually have to point this out a second time, but since apparently I do, the exact words were...
> his never ending insistance that if he were accomdated that world would
> be safe for democracy.
If that is not mocking, please point out where I said the world would be safe etc.
And never ending? Come on. Ive made a total of 12 posts in this forum to date.
I hear ya, I have an AMD
)
I hear ya, I have an AMD K6-III 400 that works perfectly fine as a web browser and word processor. It's a great little machine within the limits of what it is used it for. But, since i have other machines, it doesn't do any crunching, because to me, it's not worth spending money on the electricity to have it crunch, it just takes too long. Now if it was my only machine, that would be a different story. I'd prolly set it up to run just one project and let it go 24x7.
Obsolescence is inevitable though, the line between what's obsolete and what's viable just keeps moving forward as new stuff comes out and the line has to get drawn somewhere I guess.
> A Pentium II may not sound like that great of a processor, but the computer
> runs really well. I don't need to be told what using a fast computer is like
> because I recently bought a new Compaq computer that could probably race by
> what most of you are calling contemporary machines. The point isn't how long
> it takes for the Pentium II to process something, the point is that it is
> capable of processing it, and helping out the Einstein@Home project.