Folding Forum

bruce wrote:Change the -1 in one CPU slot to 2 like I suggested. Remove the second slot.

Ok done

Now my pc uses ~40% for folding instead of 100% before... Just wanted to fully use my cpu.

Thanks for your patient help. Going to autostart cpu folding whenever my pc start and gpu folding when the project is small.

The current work unit may only be able to use half the processor. After this work unit, the next work unit should use more CPU power if setup correctly.

7im wrote:The current work unit may only be able to use half the processor. After this work unit, the next work unit should use more CPU power if setup correctly.

running now 1 cpu project with 2 slots whenever my pc is on. gpu depends on project size

Thank you very much for your patience. I dont have any more questions. <3

7im wrote:The current work unit may only be able to use half the processor. After this work unit, the next work unit should use more CPU power if setup correctly.

When you delete a slot, the WU is moved to the other slot, so "current WU" may mean two WUs. If you set FINISH on a runnning slot, it will continue to compute, then upload, then NOT download a new WU. That would be an excellent time to delete the idle slot.

Just a suggestion: Smaller projects for older GPUs

I got a new GPU project yesterday and tried to complete it, but its again nearly impossible. My GPU need for ~0,5% one hour (full folding power). The project has a 10 day ETA. I have to fold 20 hours a day for 10 days in order to complete this one project on time.

Teddy300 wrote:Just a suggestion: Smaller projects for older GPUs

I got a new GPU project yesterday and tried to complete it, but its again nearly impossible. My GPU need for ~0,5% one hour (full folding power). The project has a 10 day ETA. I have to fold 20 hours a day for 10 days in order to complete this one project on time.

Joe_H offered this solution on the previous page.

Set the client-type to beta, and the max-packet-size to small, and the GPU will process WUs from an older project with longer deadlines.

7im wrote:

Teddy300 wrote:Just a suggestion: Smaller projects for older GPUs
Joe_H offered this solution on the previous page.

Set the client-type to beta, and the max-packet-size to small, and the GPU will process WUs from an older project with longer deadlines.

You are a bit right, but he didn´t say anything of beta client or adding values.

I maybe found it out myself. I dont know what you mean with " client-type to beta" but i added the following command in the GPU slot:
Name: max-packet-size
Value:small

Got a new GPU project with 4000 Points, the projects before were both 4300 Points. Its slightly better.

Thanks

Teddy300 wrote:Got a new GPU project with 4000 Points, the projects before were both 4300 Points. Its slightly better.

There's not much difference between a 4000 point WU and 4300 point WU (other some random factors that are unknown).

What you're reporting presented FAH with a difficult problem. There are a number of GPUs which are either too slow or which fold too few hours to complete any of the normal WUs. They needed a method got you to select projects with long deadlines -- and the Quick Return Bonus program wasn't enough to encourage people like you to fold 24x7.

They could have added an option like "log-deadlines=true" but that required building and distributing a new client. The existing option max-packet-size has another purpose, but it's rarely used, so they decided to assign those long-deadline projects to those who use BOTH max-packet-size=small and client-type=beta. That's not as clean a method as distributing a new client with a more meaningful option, but considering the cost of distributing a new client, I think it was a valid compromise.

(... especially since end-of-life has been announced for projects that use FahCore-15.)

The Configuration Guide can help with setting the client-type setting to beta same was as you set the max-packet-size setting to small before.

I have now edited my client.

Thank you all for your help, especially bruce and 7im.

I wont produce insane amounts of PDD but i will do it consistently.

Just an update:
Was folding for more then a month with my dual core and got ~40.000 points. I decided a few days ago to upgrade my graphics (gtx 960 now) card because my old was to slow. Finished my first gpu project yesterday in 9 hours and got 60.000 points

I will fold from now on with my graphics card a few hours every day until iam done with my study. Won´t give a shit about electricity then and will fold 24/7

Feels good to help the science. Keep it up

If you fold a few hours a day, you should seriously consider using the Finish setting whenever possible, thereby processing WUs to their logical completion, uploading the result, and pausing until the next time your system is scheduled to start processing again.