I think maybe when I've collected many samples, I should make Einstein two columns, one for long wus and one for short.
Yes, but in between there also differences. I've also computed the credit per hour ratio and ordered them. Then, I noticed there is a pattern. All ratios for a specific granted ratio are close to each other. So there's not only a difference between small and large, but also inside these two classes. For each of these subclasses the variation of credit per hour was very small.
I See Tony has provided some information and is getting ready for a new round with Einstein...
I have seen that one person requested a copy of "tony's" Spreadsheet that they could post data into...
I have seen a Whole Lot of Complaining and Very Few asking, "What Can I do to Help!" Bruce is concerned with two things, First that Users understand what is happening and Second making Einstein "fair" across the various BOINC Projects.
With all that Said you have information for your small world One computer or several, Intel vs AMD at various speeds... So with that said and changes to the Science Application... Changes tend to reflect in a very Large Way, Intel may have been good but AMD could be better... That is part of what accurate information will help to determine...
Looking at my machines that are doing Einstein I see a another small change has happened in the last day or so... That spreadsheet over time shows what is real and what looking in the wrong places provides as "Mis-information." I have about a dozen machines that I am moving through 5 projects... As a Result many people are passing me in Einstein, they are passing me in Seti. So while you complain, I see very Few Working to Help Make Things Better!
RE: I think maybe when I've
)
Yes, but in between there also differences. I've also computed the credit per hour ratio and ordered them. Then, I noticed there is a pattern. All ratios for a specific granted ratio are close to each other. So there's not only a difference between small and large, but also inside these two classes. For each of these subclasses the variation of credit per hour was very small.
Happy crunching,
Bert
Somnio ergo sum
I See Tony has provided some
)
I See Tony has provided some information and is getting ready for a new round with Einstein...
I have seen that one person requested a copy of "tony's" Spreadsheet that they could post data into...
I have seen a Whole Lot of Complaining and Very Few asking, "What Can I do to Help!" Bruce is concerned with two things, First that Users understand what is happening and Second making Einstein "fair" across the various BOINC Projects.
With all that Said you have information for your small world One computer or several, Intel vs AMD at various speeds... So with that said and changes to the Science Application... Changes tend to reflect in a very Large Way, Intel may have been good but AMD could be better... That is part of what accurate information will help to determine...
Looking at my machines that are doing Einstein I see a another small change has happened in the last day or so... That spreadsheet over time shows what is real and what looking in the wrong places provides as "Mis-information." I have about a dozen machines that I am moving through 5 projects... As a Result many people are passing me in Einstein, they are passing me in Seti. So while you complain, I see very Few Working to Help Make Things Better!
Have a Good Evening
Regards Pappa