Did I say anything about credit? It saves time because it does more WU's in the same amount of time and I don't have to think about upgrading the systems. That's pretty simple isn't it?
How will it save money and time to you? If you're counting wus(useless for yourself(!), just useful for the project itself) then ok. Nothing related to BOINC recuires an upgrade so you are not forced to do so. If the app gets faster and all use it your system is relatively as slow as before so if you want to speed up relative to the others the need of new hardware is clear independent of the software ...
btw: this is no attack on your person, k? I just wanted to illustrate the situation for credit hunters ... chill out :)
Puts new life back in these old cpu's and will save me both time and money.
Erm, no. Related just to credit, the new boostapp will do nothing. The only one who's getting use out of it is the project itself and science intended crunchers (like me) ;-). AFAIK simply because your credit is just referred to your crunching speed (FLOPS/IOPS) and time (crunching 10k seconds on 2 wus instead of 1 does't change the total time spent for the project, it just cuts the credit in half per wu).
Unless the other 2 people working on the same unit don't have the optimized albert app, in which case you get their credit for the WU instead of your own, and double your credit!
such things just should not be writ so please destroy this if you wish to live 'tis better in ignorance to dwell than to go screaming into the abyss worse than hell
Unless the other 2 people working on the same unit don't have the optimized albert app, in which case you get their credit for the WU instead of your own, and double your credit!
OK you're right, but it's not the aim of the project that everyone uses a different app ... I hope that all ppl will use a faster software soon
Unless the other 2 people working on the same unit don't have the optimized albert app, in which case you get their credit for the WU instead of your own, and double your credit!
OK you're right, but it's not the aim of the project that everyone uses a different app ... I hope that all ppl will use a faster software soon
Completely agreed. Maybe we should just get rid of credit to avoid half of the posts in this forum ;)
such things just should not be writ so please destroy this if you wish to live 'tis better in ignorance to dwell than to go screaming into the abyss worse than hell
How will it save money and time to you? If you're counting wus(useless for yourself(!), just useful for the project itself) then ok. Nothing related to BOINC recuires an upgrade so you are not forced to do so. If the app gets faster and all use it your system is relatively as slow as before so if you want to speed up relative to the others the need of new hardware is clear independent of the software ...
btw: this is no attack on your person, k? I just wanted to illustrate the situation for credit hunters ... chill out :)
OK, I'll make this simple for you: I have 2 P4's, one is a 2.8e and the other is a socket 423 @ 1.5Ghz, everything else is either dual or quad PIII Xeons at 500-700Mhz. Since the changes to the app allow each cpu to do more work in a give length of time, that makes the cost per WU go down. We are talking cost of the power to keep the farm running. Overall cost stays the same but the cost per WU is reduced.
Run time for the old Xeons was in the range of 75k-95k per WU, depending on the speed of the cpu in question. Time per WU has dropped to 25-30k. The only other way to increase output would be to replace the systems I already have since they are all running the fastest cpu's they will support. Or, I could add another 6-7 machines of the same speed. Hence, it saves me money because I don't need to replace the systems, or add new ones to increase the amount of work done. And, it saves me time because I don't have to spend the time, and money, looking for the parts I need nor do I have to spend the time to build new systems. It saves me time and money, with the added benefit to the project of getting more WU's done in a given length of time with the resources they have available to them.
At any time independent of the used app you could buy new hardware and increase the work done. This is not the point we are talking about. You write that the money/wu is reduced. Absolutely correct but wus are crucial for the project not for your person.
"it saves me money because I don't need to replace the systems"
-> You would save time and money if you want to reach a certain crunching speed related to wus done but this is weird. There is no "need" from the project side.
"it saves me time because I don't have to spend the time, and money, looking for the parts I need nor do I have to spend the time to build new systems"
-> you dont "have to" spend the time and money, who is forcing you to do so? Maybe you force yourself? This is independent of the software. I can force myself too to build another system, yes, now, tomorrow, yesterday ...
"I could add another 6-7 machines"
-> I could too, at any time, whether the app speeds up or not that not the point ...
Lets think about the theoretical case the S4 analysis is done and you worked until the end. If you now stop crunching anything, then you have saved money and time because you are finished earlier, ok. But I guess you will search for another task for your cpus (S5 or something completely different) and if they are finished you are again searching and so on ... so your cpus are used all the time, indepently to what project they contribute. So money and time is the same.
So we see, we all could but nobody has to upgrade. Just if you want to enlarge your contribution to the project there is a need to buy new hardware if the app doesn't change. Maybe you are so noble-minded and still want to increase working speed then it's a break for you and you would save money and time for a certain moment ... :-)
OK, let's finish with that, I think it's getting boring ^^
My Athlon XP 2800+ on Windows XP Home went from 15380 sec to 6598 sec. Claiming about half the credit, but I was rarely getting what I claimed before anyway, with the optimized BOINC client.
Time reportet is almost everywhere zero, but it took always about 13000s to complete. But luckily the workunits get validated though.
Is there anything i can do to fix the time reported?
RE: Did I say anything
)
How will it save money and time to you? If you're counting wus(useless for yourself(!), just useful for the project itself) then ok. Nothing related to BOINC recuires an upgrade so you are not forced to do so. If the app gets faster and all use it your system is relatively as slow as before so if you want to speed up relative to the others the need of new hardware is clear independent of the software ...
btw: this is no attack on your person, k? I just wanted to illustrate the situation for credit hunters ... chill out :)
Greetings, Santas little helper
RE: RE: Puts new life
)
Unless the other 2 people working on the same unit don't have the optimized albert app, in which case you get their credit for the WU instead of your own, and double your credit!
such things just should not be writ so please destroy this if you wish to live 'tis better in ignorance to dwell than to go screaming into the abyss worse than hell
RE: Unless the other 2
)
OK you're right, but it's not the aim of the project that everyone uses a different app ... I hope that all ppl will use a faster software soon
Greetings, Santas little helper
RE: RE: Unless the other
)
Completely agreed. Maybe we should just get rid of credit to avoid half of the posts in this forum ;)
such things just should not be writ so please destroy this if you wish to live 'tis better in ignorance to dwell than to go screaming into the abyss worse than hell
RE: How will it save money
)
OK, I'll make this simple for you: I have 2 P4's, one is a 2.8e and the other is a socket 423 @ 1.5Ghz, everything else is either dual or quad PIII Xeons at 500-700Mhz. Since the changes to the app allow each cpu to do more work in a give length of time, that makes the cost per WU go down. We are talking cost of the power to keep the farm running. Overall cost stays the same but the cost per WU is reduced.
Run time for the old Xeons was in the range of 75k-95k per WU, depending on the speed of the cpu in question. Time per WU has dropped to 25-30k. The only other way to increase output would be to replace the systems I already have since they are all running the fastest cpu's they will support. Or, I could add another 6-7 machines of the same speed. Hence, it saves me money because I don't need to replace the systems, or add new ones to increase the amount of work done. And, it saves me time because I don't have to spend the time, and money, looking for the parts I need nor do I have to spend the time to build new systems. It saves me time and money, with the added benefit to the project of getting more WU's done in a given length of time with the resources they have available to them.
At any time independent of
)
At any time independent of the used app you could buy new hardware and increase the work done. This is not the point we are talking about. You write that the money/wu is reduced. Absolutely correct but wus are crucial for the project not for your person.
"it saves me money because I don't need to replace the systems"
-> You would save time and money if you want to reach a certain crunching speed related to wus done but this is weird. There is no "need" from the project side.
"it saves me time because I don't have to spend the time, and money, looking for the parts I need nor do I have to spend the time to build new systems"
-> you dont "have to" spend the time and money, who is forcing you to do so? Maybe you force yourself? This is independent of the software. I can force myself too to build another system, yes, now, tomorrow, yesterday ...
"I could add another 6-7 machines"
-> I could too, at any time, whether the app speeds up or not that not the point ...
Lets think about the theoretical case the S4 analysis is done and you worked until the end. If you now stop crunching anything, then you have saved money and time because you are finished earlier, ok. But I guess you will search for another task for your cpus (S5 or something completely different) and if they are finished you are again searching and so on ... so your cpus are used all the time, indepently to what project they contribute. So money and time is the same.
So we see, we all could but nobody has to upgrade. Just if you want to enlarge your contribution to the project there is a need to buy new hardware if the app doesn't change. Maybe you are so noble-minded and still want to increase working speed then it's a break for you and you would save money and time for a certain moment ... :-)
OK, let's finish with that, I think it's getting boring ^^
Greetings, Santas little helper
Santas little helper: You
)
Santas little helper:
You have missed the point completely, three times!
for Info, with the new
)
for Info, with the new optimized Client
Albert WUs about 1:30
with my Athlon64 3.200+
Greetings from Germany NRW
Ulli
RE: for Info, with the new
)
I've tried to run akosfs code
)
I've tried to run akosfs code on a linux box by using wine, but i've got a strange effect, look at this:
http://einsteinathome.org/host/566983/tasks
Time reportet is almost everywhere zero, but it took always about 13000s to complete. But luckily the workunits get validated though.
Is there anything i can do to fix the time reported?