Overclock.net › Forums › Graphics Cards › NVIDIA › [Official] Nvidia GeForce GTX TITAN Owners' club
New Posts  All Forums:Forum Nav:

[Official] Nvidia GeForce GTX TITAN Owners' club - Page 121

post #1201 of 28134
Quote:
Originally Posted by DimmyK View Post

Quote:
Originally Posted by Mighty Customer View Post

I have an Quadro FX5800 and I`m really tempted by the computing power of this card.
below is my Cinebench 11.5 result.
I`ll be eternally grateful if some of you guys can post some Cinebench results for Titan.
Thanks in advance. Warning: Spoiler! (Click to show)

Here you go:



GPU usage is 30-50%, core clocks not boosting, OCed or not, seems Cinebench is not really optimized for Titan yet. Oh, and I tried it with double precision too, same score.

Many thanks for doing it. thumb.gif
post #1202 of 28134
Ok so it seems to throttle (drop core clock, not sure if it's actually "throttling" or not) when I have the fan speed at 85%. I think. Have to do more testing, but here's what I have so far. PT at 106%, temperature is a non-issue as this thing only hit 70C. In the pic, you can see that the last run keeps the core at 1058mhz, even though the Power % is at 90% at that moment, yet on the run before that, the core was dropping down to 1019mhz at times but Power % didn't even hit 90 headscratch.gif




And here...a comparison between first: 2 runs of the Metro bench, and then after that I ran it again (one run second time) and kept the fan on Auto instead of 85%. Look at how on the first usage chunk, where I did 2 runs, how it drops the core towards the end of each run. Then, with the fan on Auto I did one run, and you can see it's at 1058mhz the entire time, even at the end where it dropped in the 2 runs the first time. Double headscratch.gifheadscratch.gif

My Only Hobby
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 3930k Rampage IV Extreme EVGA 980Ti SC 16GB Samsung 30nm 
Hard DriveCoolingOSMonitor
Seagate 3TB 7200rpm H100 Windows 8.1 Acer XB240H 
KeyboardPowerCaseMouse
Razer Lycosa Cooler Master 1200w Silent Pro Gold Cooler Master HAF X Razer DeathAdder 
  hide details  
Reply
My Only Hobby
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 3930k Rampage IV Extreme EVGA 980Ti SC 16GB Samsung 30nm 
Hard DriveCoolingOSMonitor
Seagate 3TB 7200rpm H100 Windows 8.1 Acer XB240H 
KeyboardPowerCaseMouse
Razer Lycosa Cooler Master 1200w Silent Pro Gold Cooler Master HAF X Razer DeathAdder 
  hide details  
Reply
post #1203 of 28134
Quote:
Originally Posted by TheBlademaster01 View Post

Quote:
Originally Posted by DimmyK View Post

Here you go:





GPU usage is 30-50%, core clocks not boosting, OCed or not, seems Cinebench is not really optimized for Titan yet. Oh, and I tried it with double precision too, same score.

At those render settings you're CPU limited...

Stock 470:




The CPU is important indeed. With my Quadro FX5800 (GeForce 280 based) on a 3930K @4.3GHz I managed to score 60fps. Maybe with a similar CPU it might jump at 80-9ofps. Its not a big jump but it is an improvement. ATI cards have scores >100fps. nVidia is limiting the gaming cards to sell the Quadro versions.
Thanks for doing it.thumb.gif
post #1204 of 28134
Quote:
Originally Posted by dph314 View Post

Ok so it seems to throttle (drop core clock, not sure if it's actually "throttling" or not) when I have the fan speed at 85%. I think. Have to do more testing, but here's what I have so far. PT at 106%, temperature is a non-issue as this thing only hit 70C. In the pic, you can see that the last run keeps the core at 1058mhz, even though the Power % is at 90% at that moment, yet on the run before that, the core was dropping down to 1019mhz at times but Power % didn't even hit 90 headscratch.gifWarning: Spoiler! (Click to show)


And here...a comparison between first: 2 runs of the Metro bench, and then after that I ran it again (one run second time) and kept the fan on Auto instead of 85%. Look at how on the first usage chunk, where I did 2 runs, how it drops the core towards the end of each run. Then, with the fan on Auto I did one run, and you can see it's at 1058mhz the entire time, even at the end where it dropped in the 2 runs the first time. Double headscratch.gifheadscratch.gif


It appears the fan power draw counts against the power throttle so forcing the higher fan speed may be pushing you over the limit even if the power # on Afterburner doesn't show it. As long as temps are under control, turn the fan down, or leave it on Auto and see if that helps. Although 83% is a long way from the power limit - seems like the fan wouldn't make that big a difference. Maybe there is something else about the fan (besides power draw) that is affecting things.
Edited by Forceman - 2/28/13 at 3:46pm
post #1205 of 28134
Quote:
Originally Posted by Forceman View Post

Quote:
Originally Posted by dph314 View Post

Ok so it seems to throttle (drop core clock, not sure if it's actually "throttling" or not) when I have the fan speed at 85%. I think. Have to do more testing, but here's what I have so far. PT at 106%, temperature is a non-issue as this thing only hit 70C. In the pic, you can see that the last run keeps the core at 1058mhz, even though the Power % is at 90% at that moment, yet on the run before that, the core was dropping down to 1019mhz at times but Power % didn't even hit 90 headscratch.gifWarning: Spoiler! (Click to show)


And here...a comparison between first: 2 runs of the Metro bench, and then after that I ran it again (one run second time) and kept the fan on Auto instead of 85%. Look at how on the first usage chunk, where I did 2 runs, how it drops the core towards the end of each run. Then, with the fan on Auto I did one run, and you can see it's at 1058mhz the entire time, even at the end where it dropped in the 2 runs the first time. Double headscratch.gifheadscratch.gif
Warning: Spoiler! (Click to show)

It appears the fan power draw counts against the power throttle so forcing the higher fan speed may be pushing you over the limit even if the power # on Afterburner doesn't show it. As long as temps are under control, turn the fan down, or leave it on Auto and see if that helps. Although 83% is a long way from the power limit - seems like the fan wouldn't make that big a difference. Maybe there is something else about the fan (besides power draw) that is affecting things.

Jacob better get in here pronto. I want to know why when I turn the fan speed up on my thousand-dollar piece of hardware to keep it cooler, the core drops. Is the displayed Power % in Precision accurate? I don't think it is if I'm throttling because of the fan speed being up when I only see ~90% Power %.

Other than that, initial stock testing results are almost all in. Going to start the overclocking soon smile.gif Oh, and anyone have Titan-fever?...

My Only Hobby
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 3930k Rampage IV Extreme EVGA 980Ti SC 16GB Samsung 30nm 
Hard DriveCoolingOSMonitor
Seagate 3TB 7200rpm H100 Windows 8.1 Acer XB240H 
KeyboardPowerCaseMouse
Razer Lycosa Cooler Master 1200w Silent Pro Gold Cooler Master HAF X Razer DeathAdder 
  hide details  
Reply
My Only Hobby
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 3930k Rampage IV Extreme EVGA 980Ti SC 16GB Samsung 30nm 
Hard DriveCoolingOSMonitor
Seagate 3TB 7200rpm H100 Windows 8.1 Acer XB240H 
KeyboardPowerCaseMouse
Razer Lycosa Cooler Master 1200w Silent Pro Gold Cooler Master HAF X Razer DeathAdder 
  hide details  
Reply
post #1206 of 28134
Just ordered my second TITAN SC, it's being packed right now. wheee.gif Probably have them both by mid week. thumb.gif
post #1207 of 28134
Quote:
Originally Posted by dph314 View Post

Jacob better get in here pronto. I want to know why when I turn the fan speed up on my thousand-dollar piece of hardware to keep it cooler, the core drops. Is the displayed Power % in Precision accurate? I don't think it is if I'm throttling because of the fan speed being up when I only see ~90% Power %.

Other than that, initial stock testing results are almost all in. Going to start the overclocking soon smile.gif Oh, and anyone have Titan-fever?...

How much extra power draw is the 6GB of VRAM is causing? Wonder if you could undervolt the memory and get more headroom for the core, and if that would be better than overclocking the memory itself.

Edit: I can't make sense of the GDDR5 power info that Samsung is putting out - I see numbers that estimate the memory power draw ranging from 10W all the way to 40W.
Edited by Forceman - 2/28/13 at 4:21pm
post #1208 of 28134


Add me please
EVGA GTX Titan Superclocked.
post #1209 of 28134
Question. Still trying to fully understand this Boost 2.0.

In my situation one card is a SC, and the other a vanilla. The SC is base clocked at 876mhz and vanilla 837mhz. The boost clock for these is 928mhz and 876mhz respectively. That's fine, I'm cool with that. So then I set my GPU Clock Offset to +129 (both cards are synced up in Precision). Does that +129 add to the stock number or the boost number? I'm slightly confused here because when I run benchies/games, I'm seeing the boost clock read 1176mhz for the SC and 1137mhz for the vanilla? Shouldn't it read 1005mhz boost (SC) 966mhz boost (Vanilla)? Power is at 106% and temp at 90c by the way.

Sorry for all of the numbers, hopefully someone can follow.
Edited by dboythagr8 - 2/28/13 at 4:46pm
post #1210 of 28134
Quote:
Originally Posted by dph314 View Post


Other than that, initial stock testing results are almost all in. Going to start the overclocking soon smile.gif Oh, and anyone have Titan-fever?...


buttkick.gif
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: NVIDIA
Overclock.net › Forums › Graphics Cards › NVIDIA › [Official] Nvidia GeForce GTX TITAN Owners' club