I just read the whole thread, I haven't yet tried this app, but I have one major concern: If this is an app that was written outside E@H, and differs so massively in calculation time, how can one be sure that it delivers the right scientific results? I'm here for the science and I would not want to deliver wrong results and skew the scientific outcome because of an urge to calculate faster.
because it validates against the stock application. If it was generating significantly different results, it would get mostly invalid results and no credit. Petri did a lot of internal testing to make sure it validated well.
I just read the whole thread, I haven't yet tried this app, but I have one major concern: If this is an app that was written outside E@H, and differs so massively in calculation time, how can one be sure that it delivers the right scientific results? I'm here for the science and I would not want to deliver wrong results and skew the scientific outcome because of an urge to calculate faster.
You won't have to worry about that issue. Keith has been a 'User of the Day' at Einstein while using the app. I doubt very much if his calculations were off and still been a UoD.
You won't have to worry about that issue. Keith has been a 'User of the Day' at Einstein while using the app. I doubt very much if his calculations were off and still been a UoD.
Ha ha LOL. You can have no credit at all on projects and still be selected UoD. It just means you have an account and have been selected randomly by luck of the draw.
You won't have to worry about that issue. Keith has been a 'User of the Day' at Einstein while using the app. I doubt very much if his calculations were off and still been a UoD.
Ha ha LOL. You can have no credit at all on projects and still be selected UoD. It just means you have an account and have been selected randomly by luck of the draw.
Shhh!! .....you should have played it out. I didn't know that, and I couldn't have said what Ian did because I didn't know enough. :>)
Great question, and I believe you may rest easy. A very small percentage of E@H users will use this special app; my guess is at least 99.9% of users are running stock Linux or Windows apps. For the very small percentage of users with the special app, the results will be compared to results of wingmen (remember, it takes 2 users with the same results to validate); that the special app's results validate successfully against folks running the stock apps should answer "how can we be sure?" The odds of a user running the special app having a wingman running the same special app is very remote IMHO.
In random review of my completed and validated results, I haven't found a case yet when it validated with someone else running the custom app.
Apologies for the redundant post. I saw the question several hours earlier and typed up most of the reply; got called away to help with Dinner and didn't re-load the thread prior to finishing-up and posting. Should have known the A-team would be all over it.
Happy to report even the 1660 Ti (a) can run V1.0 and (b) sees improvement over 0.95 (all running just one task at a time). The ~450 sec times with stock app shortened to 355 seconds with V0.95, and if otherwise idle, V1.0 tasks clock in at 305 seconds. In all, going from stock to V1.0 is just a hair short of a time reduction of a third, or put another way, allows completing almost 50% more tasks in a period of time.
Wow. Santa was indeed good. Now, to just not run out of work units for which this benefits any time too soon...
Is it "new" never been processed data? Or "new" let's take a look at data that was processed with older apps?
Who knows. :)
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
Looks like we're chewing through WUs faster than the server stats algorithm was predicting. Just under 3 weeks ago it estimated 341 days of work left. Today's guesstimate is 241 days.
At least I should get more than the 2 months I got on Petri's SETI app before it closed (time starting with purchase of first compatible card).
stfn wrote: I just read the
)
because it validates against the stock application. If it was generating significantly different results, it would get mostly invalid results and no credit. Petri did a lot of internal testing to make sure it validated well.
_________________________________________________________________________
stfn wrote: I just read the
)
You won't have to worry about that issue. Keith has been a 'User of the Day' at Einstein while using the app. I doubt very much if his calculations were off and still been a UoD.
Proud member of the Old Farts Association
GWGeorge007 wrote: You won't
)
Ha ha LOL. You can have no credit at all on projects and still be selected UoD. It just means you have an account and have been selected randomly by luck of the draw.
Keith Myers
)
Shhh!! .....you should have played it out. I didn't know that, and I couldn't have said what Ian did because I didn't know enough. :>)
Proud member of the Old Farts Association
Great question, and I believe
)
Great question, and I believe you may rest easy. A very small percentage of E@H users will use this special app; my guess is at least 99.9% of users are running stock Linux or Windows apps. For the very small percentage of users with the special app, the results will be compared to results of wingmen (remember, it takes 2 users with the same results to validate); that the special app's results validate successfully against folks running the stock apps should answer "how can we be sure?" The odds of a user running the special app having a wingman running the same special app is very remote IMHO.
In random review of my completed and validated results, I haven't found a case yet when it validated with someone else running the custom app.
Regards,
Fred
Apologies for the redundant
)
Apologies for the redundant post. I saw the question several hours earlier and typed up most of the reply; got called away to help with Dinner and didn't re-load the thread prior to finishing-up and posting. Should have known the A-team would be all over it.
Happy to report even the 1660
)
Happy to report even the 1660 Ti (a) can run V1.0 and (b) sees improvement over 0.95 (all running just one task at a time). The ~450 sec times with stock app shortened to 355 seconds with V0.95, and if otherwise idle, V1.0 tasks clock in at 305 seconds. In all, going from stock to V1.0 is just a hair short of a time reduction of a third, or put another way, allows completing almost 50% more tasks in a period of time.
Wow. Santa was indeed good. Now, to just not run out of work units for which this benefits any time too soon...
After threatening that
)
After threatening that FGRPB1G was going to end soon, Bernd elf-magicked a boatload of new work.
FGRPB1G search progress
Is it "new" never been
)
Is it "new" never been processed data? Or "new" let's take a look at data that was processed with older apps?
Who knows. :)
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
Looks like we're chewing
)
Looks like we're chewing through WUs faster than the server stats algorithm was predicting. Just under 3 weeks ago it estimated 341 days of work left. Today's guesstimate is 241 days.
At least I should get more than the 2 months I got on Petri's SETI app before it closed (time starting with purchase of first compatible card).
FGRPB1G search progress