• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Unigine Heaven 4 benchmark

The strong men of the 2 and 4 GPU tables have arrived, they will be holding all the others up.:D

Single HD 5970 stock 725/1000
CPU i7 975 @3.47

4qta.jpg


2 x HD 5970s quadfire stock 725/1000
CPU i7 975 @3.47

2fk.jpg


Does anyone know where I can get a copy of the XFX overvolt tool so I can overclock these again (I have lost my copy).
 
RetrogamerX said:
I guess the ATI tool worked then kaap?
Click to expand...

I have not got to use it yet, with hindsight I wish I had. Someone else said about giving MSI afterburner a go, which is what I used here. Unfortunately when I had finished and when back to stock settings something went wrong and my installation of windows got trashed. I spend several hours last night sorting it out.

Next time I think I will give the ATI tool a go instead.:)
 
Guru said:
Wow!!

It's weird how you beat the person just below you despite him having higher clocks too! :)
Click to expand...

The clocks on his GPUs are the boost clocks, mine are the un boosted ones. With the NVidia 6 and 7 series cards the clock speeds can get very confusing (for me anyway).:)

Just ran this to show the difference between set and boost clocks. In the pic below I have used the clocks I used in the Heaven 4 bench, with this I am running the first test in 3dmark11 (it is easy to setup and run). I am only running my CPU @4.0 as I am not after a big score.

You can see the maximum boost the GPUs are reaching and the set clocks.

e2ik.jpg
 
Last edited:
Guru said:
Ah I see! I'm considering a 780 if I sell my 2x 570's
Click to expand...

GTX 780s are great cards, check out the scores on the scoreboards.:)

I keep getting the urge to get a GTX 780/i7 4770k system as you can have a lot of fun benching them.:D

I have added a pic in my previous post showing the difference between set and boost clocks on my cards, this also works the same with GTX 780s.
 
Duff-Man said:
Niiice :) Truly king of the multi-GPU tables now.

I see that you haven't put the water chiller into your loop yet? Still waiting for cooler weather?


Just so you know, there is no way for the temps to drop sub-ambient, unless the chiller is capable of handling the entire wattage of your loop... The closer you get to ambient, the less effective the radiator will be. The loop will find a new equilibrium temperature, which will be closer to ambient, but will not be below it unless the chiller can handle the entire heat output from the loop. At this point you would be better off removing the rad from the loop, as it will only serve to heat the water.

Apologies if you already knew all the above... What is the heat removal rating for your chiller?
Click to expand...

Thanks.:D:)

It is a Hailea HC-500A

I don't think it will be able to replace the rads as my rig produces way too much heat for that. What I want to do is use it to work with the rads to knock a few degrees of the temps. I plan to fit it after the last rad and before the CPU as it is the CPU that needs the most help. If I can knock 5 or more degrees off the temps on a cold day it will be worth it, I know using the rads as well I will be limited to above ambient temps. The other thing I want to avoid is condensation, I don't want to use insulation on the parts so going too low on the temps is a no no.

Having said all that though, for a single card and CPU the chiller will be able to cope on it's own so the option is there to turn off the fans on the rads and do some serious benching on a single card.:D
 
Last edited:
Duff-Man said:
Hmm - well I'd be very interested to see the results when you put it in the loop.

Thermodynamically speaking, the chiller will effectively "cancel out" some of the heat generation in the loop. So, if the components in your loop produce X Watts, and the chiller is rated to remove Y Watts, then the water temperature in the loop will be the same as if you were running (X-Y) Watts without the chiller. So if your chiller can remove 250W (say), the water temperature in your quad-GPU setup will be very similar to what you get with a tri-GPU setup minus the chiller.



Hmm... seems that chiller is rated to remove up to 780W(!!!), though that might change somewhat depending on flowrate. So, under load, your performance with a quad-Titan system would be pretty similar to what you get with a single-Titan in the same loop - minus the chiller.

The concern would come when you're not under load since you'll certainly be drawing less than 700W from the loop. The radiators would provide a buffer, acting to warm the water back towards ambient, but condensation could be a concern. If I were you, I'd be tempted to add the chiller, let it drop to one or two degrees below ambient, and then run the benchmarks. The volume of water will take a while to heat up, so you could essentially run benchmarks at ambient, or just below. When your water temps rise, just drop back to stock for a while until you reach your target temperature :)

Sod waiting for winter - just pick a relatively cool day, bench at night when the air temperature has dropped to 18 degrees or so, open a window, and get cracking :D
Click to expand...

I do have a problem maybe you could help with. When the chiller arrived there was a little bit of water at the bottom (probably what they used to test it with). Do you have any ideas how I can get it out, from the manual it says I must not turn the chiller upside down. The only way I can think of is to use the same type coolant I use in the loop to purge it out.

Thanks.
 
You must log in or register to reply here.
Back
Top Bottom