That's Linux, it's doing more work while requiring less cpu usage. I think you're running a1 core wu's btw then as they always used about 40% here per vm but the a2's were abit higher. Still had enough to even run mj12 next to gpu folding.Originally Posted by p2501
Marvin, how did you set your priorities to make everything work so well together?
I set the GPUv2 priority to normal manually and that of both VM's to idle in task manager, still I get a points hit on the first GPU...
Edit: I think that's solved, I'm using ProcessLasso to assign priorities and cores, both GPU's back to normal. You said you have two SMP VMX instances running, did you set both up to use 4 or 2 cores (not cores affinities, but in the setup of the VM's)? I'm just asking bc the cores are not 100% loaded, they are varying around 92-100%.
I set both to use 2 cores, but even if you said four it wouldn't matter because the vm only exposes dual cores. What I should do is use that program Afc and set up each vm to use two cores on the same die ( which atm I don't know if it's two on one die or two spread over two dies causing heavy fsb usage. If rr or you pick up a 2669 we can compare ppd abit and if any of us have considerable higher or lower points the core logic is a logical thing to look at ( pun)
Btw if you set gpu2 priority manual in taskman it will reset itself on the next work unit ( you did change the fahcore11 or fahcore13 right? not the client itself? ). I set the priority to 'low' in the config options, still got abit of a point hit but not much and most of it is due to mj12.
Network\FOLD-%random string%\etc\folding\1 ( for first instance if set to smp per 2 cores and you run it on a quad core ) or \2 for second instance.
Enable network discovery if you didn't have that on, and during vm setup be sure to leave windows networking enabeld
Bookmarks