S5R4 takes 50 % more time to crunch but the value is instead of 237 points only 62 points, something must be very wrong or the client or the value
workunit ID 42220541
Result ID 103506183
S5R4 takes 50 % more time to crunch but the value is instead of 237 points only 62 points, something must be very wrong or the client or the value
workunit ID 42220541
Result ID 103506183
Hi!
Please see this message from Bernd. The "claimed credit" is not important for Einstein@Home, the credit that you will finally be awarded will be much higher.
S5R4 takes 50 % more time to crunch but the value is instead of 237 points only 62 points, something must be very wrong or the client or the value
workunit ID 42220541
Result ID 103506183
Computer sind nicht alles im Leben. (Kleiner Scherz)
Thank you for your fast reply, so if i understand you correctly the awarded credit should be at least a factor 5 times higher then the claimed if the same relation crunching time versus awarded credit is to be maintained as in S5R3
Thank you for your fast reply, so if i understand you correctly the awarded credit should be at least a factor 5 times higher then the claimed if the same relation crunching time versus awarded credit is to be maintained as in S5R3
It's more like ca 3 times the claimed credit, but on the other hand the extra runtime is far less than 50% more.
A slight credit reduction in comparison to S5R3 is intended to bring E@H back in line with other major BOINC projects like SETI. It's somwhat "unfair" (or let's say, unkind) to award much higher amounts of credits for the same CPU time spent.
...It's somwhat "unfair" (or let's say, unkind) to award much higher amounts of credits for the same CPU time spent. Bikeman
As with beauty, fair or unfair, kind or unkind, is a matter of opinion. Credit fairness is a hot topic on most message boards and also a matter of opinion. Reality is, Bruce Allen et.al. will decide what is fair to grant; volunteers will decide with their feet what is unfair.
Thank you for your fast reply, so if i understand you correctly the awarded credit should be at least a factor 5 times higher then the claimed if the same relation crunching time versus awarded credit is to be maintained as in S5R3
It's more like ca 3 times the claimed credit, but on the other hand the extra runtime is far less than 50% more.
A slight credit reduction in comparison to S5R3 is intended to bring E@H back in line with other major BOINC projects like SETI. It's somwhat "unfair" (or let's say, unkind) to award much higher amounts of credits for the same CPU time spent.
CU
Bikeman
Dear Bikeman
My average in S5R3 was around 15.316 sec. and this was awarded with 237 credits, now i need 24.518.85 sec. which, if awarded same should be 382 credits. As in the first crunched examples visible the new awarded credits are around 160 credits instead of 382 which is less then 50% . This leads to the conclusion that it is no longer desireable to crunch for Einstein because obviously we get cheated for more then 50% of the credits.
I only crunch about 12 work units per day but you will miss them in future and i think if not changed the credit politics more will follow.
have a nice day.
What really makes me wonder is this: if you look at this result list, the S5R4 results do not take that much longer than the S5R3 results you can see still in the list, but the units I got on my own hosts tend to be about twice as long the average S5R3 units.
So I guess it's "runtime variation analysis time" again :-). Not all S5R4 units are created equal, some will take longer than others.
The credits (claimed and awarded) seem to take this into account to *some* degree (I've seen claimed credits in the 50s and the 60s, and granted credit is usually almost three times the claimed credit).
To sum it up: let's not jump to conclusions based on single data points.
validaton error with new client
)
second workunit same game62.88 points for more than 50 % longer crunching
Workunit ID 42220544
Result ID 103506183
Should i abort the rest of the downloaded Workunits until the error is fixed?
RE: S5R4 takes 50 % more
)
Hi!
Please see this message from Bernd. The "claimed credit" is not important for Einstein@Home, the credit that you will finally be awarded will be much higher.
CU
Bikeman
Thanks Bikeman I was about
)
Thanks Bikeman
I was about to reply with exactly the same link.
Well done to the team on a monumental task over the weekend.
Warped.
Warped
RE: S5R4 takes 50 % more
)
Computer sind nicht alles im Leben. (Kleiner Scherz)
Thank you for your fast
)
Thank you for your fast reply, so if i understand you correctly the awarded credit should be at least a factor 5 times higher then the claimed if the same relation crunching time versus awarded credit is to be maintained as in S5R3
RE: Thank you for your fast
)
It's more like ca 3 times the claimed credit, but on the other hand the extra runtime is far less than 50% more.
see this list of results.
A slight credit reduction in comparison to S5R3 is intended to bring E@H back in line with other major BOINC projects like SETI. It's somwhat "unfair" (or let's say, unkind) to award much higher amounts of credits for the same CPU time spent.
CU
Bikeman
RE: ...It's somwhat
)
As with beauty, fair or unfair, kind or unkind, is a matter of opinion. Credit fairness is a hot topic on most message boards and also a matter of opinion. Reality is, Bruce Allen et.al. will decide what is fair to grant; volunteers will decide with their feet what is unfair.
RE: RE: Thank you for
)
Dear Bikeman
My average in S5R3 was around 15.316 sec. and this was awarded with 237 credits, now i need 24.518.85 sec. which, if awarded same should be 382 credits. As in the first crunched examples visible the new awarded credits are around 160 credits instead of 382 which is less then 50% . This leads to the conclusion that it is no longer desireable to crunch for Einstein because obviously we get cheated for more then 50% of the credits.
I only crunch about 12 work units per day but you will miss them in future and i think if not changed the credit politics more will follow.
have a nice day.
Never mind
)
Never mind
Hi all! What really makes
)
Hi all!
What really makes me wonder is this: if you look at this result list, the S5R4 results do not take that much longer than the S5R3 results you can see still in the list, but the units I got on my own hosts tend to be about twice as long the average S5R3 units.
So I guess it's "runtime variation analysis time" again :-). Not all S5R4 units are created equal, some will take longer than others.
The credits (claimed and awarded) seem to take this into account to *some* degree (I've seen claimed credits in the 50s and the 60s, and granted credit is usually almost three times the claimed credit).
To sum it up: let's not jump to conclusions based on single data points.
CU
Bikeman