| |||||||
| nVidia/ATI GPU2 clients Folding clients that use GPUs on graphics cards running the GPU2 client |
 |
| | LinkBack | Thread Tools | Rate Thread |
5th November, 2008, 04:56 PM
| ||||
| ||||
| XP nvidia performance quirk I am interested in some feedback. I have 2 machines that are similar, both on XP, running dual cores. Machine 1, using 9600GSO Machine 2, using 9800GT. When I configured #1 with GPU folding, I noticed that the cpu core it uses fluctuates, and typically shows idling ~20%. This is different than my ATI system, and it is getting over 5k PPD. I am also running a single client in service mode. When I set up #2 with the 9800GT, everything is similiar to #1, but PPD was less than 4K. I went to the single client, and affinity shows both cores checked. I set it to core 0 only. GPU usage of core 1 increased to 50% (max), with no idling, and my PPD shot up to almost 6K PPD. Because #1 is running CPU-FAH as a service, XP will not let me touch affinity for it. I'll mess with it later to see if I get a boost, but am doubtful. EDIT: Links to free affinity setting apps would be appreciated. I am curious to see why my machines behave the way they do... I may mess with my ATI rig as well, even though it pegs core1 to max already.
__________________  Last edited by SteveI; 5th November, 2008 at 04:57 PM.      |
|
5th November, 2008, 05:24 PM
| ||||
| ||||
| Behavior is definitely different. ON #1, I manually ran the single client, (off service). The gpu took a full 50% of cpu core, and the single client took the other 50%.... no change in performance tho. This is with affinity set to core 0 or both on the single client. On #2, I have to force the single to core 0, and the gpu needs the full 50%, in order to fold at full potential.
__________________ Last edited by SteveI; 5th November, 2008 at 05:24 PM. |
|
6th November, 2008, 07:09 AM
| ||||
| ||||
| Quote:
Or use Vista. As I have said before it handles the affinities differently and is supposedly less troublesome in this regard [but that isn't enough for me to switch as yet].
__________________ #1: Tt Armor, ASUS Maximus Extreme, QX9650@4.1G, 8G Corsair Dominator GT DDR3-2000, Corsair HX1050, H2O-Swiftech, Gigabyte GTX470/Arctic Accelero Xtreme Plus II, Intel 520 SSD, Kingston SSD, 2xRaptor 150G RAID0, Win 7 Pro 64 #2: Tt Shark, ASUS P5Q Pro Turbo, Q6600@3.8G, 4G HyperX-1600, Corsair HX850, CoolerMaster V10, 2xASUS 9600GT, 2xRaptor 74G RAID0, OCZ Vertex 4 SSD, Gentoo/siduction Linux [64-bit] #3, #4: Opteron 170@2.75G nude, A8N-SLI Deluxe, Gentoo  Our sister site: www.gamersonlinux.com |
|
14th November, 2008, 03:56 PM
| ||||
| ||||
| I have a similar experience. #1 Asus socket 939 w/2.0 single core, 9600GT- 4.2K ppd #2 Asus socket 939 w/2.0 dual core,9800GT- 5.9K ppd not sure why the big difference. still looking. Ron
__________________ Folding 24/7 with AMD QuadCore CPU's & nVidia Cuda Powered GPU's  |
|
| Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
| Thread Tools | |
| Rate This Thread | |
| |
Similar Threads | ||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| AMD takes performance crown from NVIDIA with new 4870 X2 | Daniel ~ | Graphics and Sound cards; Speakers and other Peripherals | 1 | 14th August, 2008 06:32 PM |
| Nvidia Reveals Tesla Solutions for High-Performance Computing. | Samuknow | General Hardware Discussion | 5 | 25th June, 2007 07:37 PM |
| NVIDIA nTune 2.00.17.05 Performance Optimization and Monitoring for nForce mainboards | aJ\/Tg | EPoX MotherBoards | 23 | 30th October, 2004 07:06 AM |
| BIOS quirk with 8K9A | Switch | EPoX MotherBoards | 1 | 2nd December, 2002 05:23 AM |
| odd powerstrip quirk | drow_elf | Graphics and Sound cards; Speakers and other Peripherals | 0 | 2nd May, 2002 05:21 AM |
