• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Is my PPD low for 560 Ti?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
capttripppp

capttripppp

Member
​
Joined
Apr 29, 2011
Location
Anchorage, Alaska
A bought a new 560 Ti Hawk, it has 1900 on the shaders stock. Right now I am running it @ 2035 on the shaders, but I am only pulling down like 13,500 for PPD. Shouldn't this be higher? I mean crap, I'm running like 13,690 PPD on my GTX 465. I thought I was going to raise the bar a bit with this purchase. Oh, also I am running the latest drivers. Could that be the problem? I have read a few places that people had to roll back drivers for some 560 Ti's, but I don't think that they were the same cards as this. Any thoughts? Thanks
 
PPD varies from WU to WU, are you comparing them on the same WU? Are you perhaps running the 560 at idle priority (should be low)?
 
ChasR said:
PPD varies from WU to WU, are you comparing them on the same WU? Are you perhaps running the 560 at idle priority (should be low)?
Click to expand...
I did have it only idle, but I just changed it so I will see if it makes a difference. And both cards were running 6804 with different clones, so that should be close enough to compare, yes? Either way the addition will up my ppd to 26000, but I was hoping to get somewhere in the neighborhood of 15-17k ppd with the 560 Ti. Is that a realistic expectation?
 
I would expect the 560 to outproduce the 465 by a couple of thousand ppd on the same project. If you aren't folding SMP on the CPU, priority won't make much if any difference. If you are be sure the SMP client is set to idle priority. Can you compare the GPU utilization using Afterburner or Precision. It should stay pegged at 99% on most WUs. If running Windows 7, do you have the nVidia control panel set to prefer maximum performance?
 
I haven't been running an smp client. Using the Phenom II that I have didn't get me alot of ppd, so I am choosing instead to set up a gpu based rig for folding. Nvidia control panel is set to max performance, and afterburner is showing the cores of both pegged @ 99%. I reckon I will tinker with it. Maybe try rolling back the drivers to something later. I don't know what else to try. I'm just happy to be back up folding, been down for a few days because of a move
 
Just to give you an idea, my 570@825 averaged about 17k PPD across several different types of WUs. One of the 460s@950 averaged between 12-13k. I haven't folded on the 560ti yet, but based those two nvidia GPUs that are basically on either side of a 560ti performancewise, I'd imagine it being around 15k OC'd at 950, which is stock OC for a Hawk 560ti. They were all running on GPU Tracker V2.

I'm not running any GPUs right now due to high power bills, saving that up for the Chimp Challenge
 
Last edited:
THere haven't been any reports of performance differences between drivers that I have seen, just that some of the newer drivers don't work with FAH on some cards. What are you monitoring the clients with?
 
Monitoring with HFM. Tomorrow i will try to install a new client onto the 560Ti when I set it up I moved my 465 over and just set up a new client for it, and used the old client for the 506 Ti. That's really the only thing that I can think of,
 
Clients make little difference. FahCores do the work. The only time a client makes a difference is when it will run a FahCore that another client wont.
 
I was wondering you any of you thought that maybe ver7 would have any effect on the ppd level. It was just a thought, and today I have some time that I can use to play with it and figure something out. Or I am prepared to just ride out the numbers that I pulling down now if it's necessary
 
Clients don't make a difference, FAHcores do. If you aren't making the expected ppd, it's most likely that the card isn't running as fast as you think or something you are doing is vying for GPU or cpu resources needed to optimize folding production. Post up some benchmark numbers from HFM on both the 560TI and the 465. If there is a significant difference in the minimum frame time and the average frame time, something is slowing down the 560.

Data from a 8800 GTS running GPU3:
Project ID: 11177
Core: OPENMMGPU
Credit: 1349
Frames: 100


Name: Marie 8800 GTS
Path: \\Abco-cahill\FAH\GPU\
Number of Frames Observed: 300

Min. Time / Frame : 00:03:16 - 5,947 PPD
Avg. Time / Frame : 00:03:19 - 5,857 PPD


Notice there is very little difference between the min and average frame time.

Posting the benchmark data from a currently running WU would show conclusively if the 560 were slowing down or running constantly slow. Dropping to 2D clocks has been a problem with some drivers on some cards.

Cuda had a 560 TI making 17,000 ppd at about the speed you say you are running.
 
Cattripp,
Hardware Canucks tested a 560 Ti at stock clocks, 822/1644 and got 14,350 ppd on p6805. Scaling losses on GPUs are about 20% so your 1000+ MHz card should be producing close to 17,000 ppd on that and similar WUs (21.65% increase in clock rate x.8 = 16.8% increase in ppd). If you're not making more than a 560 Ti at 822 Mhz, either the card isn't running at the clocks you have set, is throttling, or you're using it for something other than folding.
 
Looking at the benches I guess it running alright. I thought I would be able to get more out of it. Maybe I will with the next one I buy. Gonna sell the 465 here soon then get another 560Ti. I'm thinking about going with Asus next time around
First is the 560, and second is the 465
 

Attachments

  • 560 Ti.PNG
    560 Ti.PNG
    19.1 KB · Views: 228
  • 465.PNG
    465.PNG
    19.3 KB · Views: 227
Which is which? gpu2 is struggling and looks like a card that has dropped to 2d clocks.
 
I see first is the 560 which is running normally and consistently, gpu2 is the 465 which isn't running right.
 
You must log in or register to reply here.

Similar threads

dfonda
Good news, Bad news. ('Hawk gets reprieve)
Replies
2
Views
179
dfonda
dfonda
don256us
#1 ppd is held by a new person. Will there be an answer?
Replies
4
Views
420
Farwalker2u
Farwalker2u
ihrsetrdr
More CUDA cores = more PPD...yes?
2
Replies
26
Views
1K
ihrsetrdr
ihrsetrdr
Farwalker2u
What is normal variance in PPD?
Replies
10
Views
610
FoldingAddict
FoldingAddict
don256us
Team challenge available, 5 billion and a weekly update too.
2
Replies
22
Views
615
HayesK
H
Back