So, I had lunch with my friend with the 6 core. He said, and now I understand, he is only running WCG, and WCG is very slick and efficient with its scheduling, so he is staying with work.

I, on the other hand, have a mixed bag of projects, and possibly, as alluded to or flat out stated above, BOINC just does not do as good a job mixing and matching, while WCG has no such problems with which to deal.

So, if that is an accurate assessment, then I am fine. If not, then please let me know.

I have decided to put my 2 desktop computers on 100%/100% on the CPU, and made the appropriate local pref changes and WCG profile changes.

Thanks to all, happy crunching.

So, first, my friend does not have 2 days additional. He has
Additional work = 0. Does that mean 6X24X0=0 hrs? I do not understand yourarithmetic.

Second, wcg cache is not my question. Overall work done is my question, I think well answered by the scheduling issue, and, the difference is well worth it so that I can contribute to a multiplicity of projects.

And, now, on my 4 core machine, I will be at 100% duty cycle. So I will be at 27.43 hours of work? That's per project, a total of 192 hours of work.

But, I still do not get your arithmetic.

Ta-Ta

I also do not understand your question it seems, what I understand, and am trying to help you understand is: my friends machine is faster than mine, but he has 40 jobs pending and I only have 1 or 2, why is he not keeping up? am I close?

the default cache is .3 days + connect every x days setting (I don't remember default). so if connect every is set to .25 days and additional work is set to zero, where I used 2 days, place .25 days. this is to insure that boinc does not run out of work before next fetch request.

the number of jobs is not inportant, the estimated completion time is, if he has mostly HCC jobs at 2 hrs each, vs your FAAH at 10 hrs each, with the same size cache, he will have 5 times as many jobs.

lastly It seems to be you are confusing cache(pending) with backlog, the faster machine will have more jobs pending, to ensure that it does not run out. this does not mean it is not completing as much, just the reverse, it means the manager thinks it will get more done.

Hope this helps, if I have misunderstood, please explain where and I will adjust my thinking and try again

fredski-

I think this is done. my friend only crunches for WCG. So his scheduling is smoother and simpler. I crunch for WCG and a bunch of other BOINC projects, which apparently creates scheduling complexities.

I have seemingly since December overcome the shortcomings of BOINC scheduling complexities by adding Additional work = 2.0 days.

This is how it was explained to me by my very erudite friend with the 6 core machine.

I think this is done.

Sek-

Yes, I will go back and enter resolved.

I have long since weighted WCG at 200 in all profiles and on all five machines. I have been extremely satisfed with what this weighting provides.

I remember that when WCG originally went to the BOINC agent, I felt that it was being treated unfairly when considered as one project. I did my weighting of 200 very soon after the advent of BOINC for WCG.

I see a lot of work from WCG. In fact you can see from my BOINC Stats bloc, WCG is the highest in BOINC credits, and I was around for UD.

So, yes, I am happy at 200, and yes, this question is resolved, as long as no one disagrees with what my friend related regarding BOINC scheduling.