• 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

spikerules said:
Ah ok, I'd heard that but wasn't sure. Thanks! Been trying out Crysis 3 and I'm not getting good results, so I was getting a tad worried.
Click to expand...

I own a pair of GTX 590s, on the mobo they are using one runs @x16 and the other @x8. When I change this so both run @x16 and run an older version of heaven, the difference in framerates is less than 0.5 of a fps. Total performance difference is also less than half of one percent.
 
How is boost calculated?

My 670 shows 1100 in GPUz but actually boosts to 1202. Thats another +102 on the core!

If I set memory clock or power higher then the boost doesnt go as high. I think the memory chips are standard but the core is a reasonably good one
 
How much a GPU boosts is connected to how hot it is running. Once a 6 series card hits 70c it will boost 13mhz lower than if it was running @69c or lower. If the GPU reaches 80c it will lose another 13mhz in boost and another 13mhz at 90c. So the trick is to keep the card as cool as possible.

If you set the power higher the card will be able to run higher clocks but this can be cancelled out by the GPU running hotter and losing boost.

One way to monitor boost is to watch the GPU clock speed on the sensor tab of GPUz
 
Kaapstad said:
How much a GPU boosts is connected to how hot it is running. Once a 6 series card hits 70c it will boost 13mhz lower than if it was running @69c or lower. If the GPU reaches 80c it will lose another 13mhz in boost and another 13mhz at 90c. So the trick is to keep the card as cool as possible.

If you set the power higher the card will be able to run higher clocks but this can be cancelled out by the GPU running hotter and losing boost.

One way to monitor boost is to watch the GPU clock speed on the sensor tab of GPUz
Click to expand...

Thanks, is there seperate memory and core temps? I only have one and have a custom fan curve so it has never goes above 67c.

Seems weird to me that I set boost to +120 so in theory should hit 1100 max. But instead boosts to 1202 (+222)

60c - 65% fan
65c - 70% fan

Edit - all explained here - http://www.overclock.net/t/1265110/the-gtx-670-overclocking-master-guide

My Kelper boost is 104. So 980+120+104 = 1204

The Max Boost is found by adding the max Kepler Boost to the Boost clock (Max Kepler Boost + Boost Clock = Max Boost). This is the most significant number when talking about the overclock on your Kepler-based GPU. It's the ONLY number alone that gives us any indication of the performance of your GPU. So please, when describing your overclock, use THIS number, not the offset, not the Kepler Boost, not the Boost Clock, and certainly not the GPU clock
Click to expand...
 
Last edited:
980MHz/6008Mhz GPU, [email protected] Turbo Boost CPU, Air cooling
heaven-1.jpg


1920x1200 = 83,9 FPS
1920x1080 = 91,6 FPS

So there's quite a bit difference between those two resolutions !
 
Last edited:
Guest2 said:
Thanks, is there seperate memory and core temps? I only have one and have a custom fan curve so it has never goes above 67c.

Seems weird to me that I set boost to +120 so in theory should hit 1100 max. But instead boosts to 1202 (+222)

60c - 65% fan
65c - 70% fan

Edit - all explained here - http://www.overclock.net/t/1265110/the-gtx-670-overclocking-master-guide

My Kelper boost is 104. So 980+120+104 = 1204
Click to expand...

The main thing is keeping your temps below 70c then you will get the maximum boost available.
 
Kaapstad said:
The main thing is keeping your temps below 70c then you will get the maximum boost available.
Click to expand...

not really, If I up the power limit % from 118 to 120 or 122 then it fails. Upping core voltage doesn't seem to help either

It would be simple if it was a case of 'carry on overclocking until temps reach 70' ...but its not
 
Guest2 said:
not really, If I up the power limit % from 118 to 120 or 122 then it fails. Upping core voltage doesn't seem to help either

It would be simple if it was a case of 'carry on overclocking until temps reach 70' ...but its not
Click to expand...

Thats done to how good your GPU is, rather than how the boost works.

On my 690s if I set the power to 135% the maximum boost available goes down as the GPUs are getting hot. If I set the power to 120% I get higher boost as the GPU is not getting so hot.

Where life gets really complicated is on benchmarks if I set the power to 135% I get a bigger score even though the max boost is lower. I think the higher power helps with the min/average fps.

The main thing is if you keep your temps below 70c this will help you with whatever other settings you are using.
 
Think I have found the 'sweet spot' on my 670

+110 core
+118 power
+550 memory

921 in heaven 4

Before it would boost to 1202 but now only to 1192 (-10 on the core after Kelper boost) but a 250mhz increase on the momory (+300 > +550)
 
Kaapstad said:
If you could do your results shot over the cobbles please, I will put you on the scoreboard.

:)
Click to expand...

Hi Kap

I thought you might say that, 3 bloody times I did this and used print screen and It pasted a black screen in paint, so finally I left the benchmark running until I got a good capture, hence it ran on past the cobbles..any thoughts ?
 
JAKUS said:
Hi Kap

I thought you might say that, 3 bloody times I did this and used print screen and It pasted a black screen in paint, so finally I left the benchmark running until I got a good capture, hence it ran on past the cobbles..any thoughts ?
Click to expand...

I am afraid we need the cobbles or the other posters won't be very happy with me.

Have you tried using FRAPS that does a good screenshot.
 
Kaapstad said:
I am afraid we need the cobbles or the other posters won't be very happy with me.

Have you tried using FRAPS that does a good screenshot.
Click to expand...

Lol Only because I beat you :D , If I get a chance I will try fraps, what was the deal with the cobbles anyway ? I do actually remember something about it before but my take is the longer it runs after the bench proves it's stable :D
 
Kaapstad said:
I am afraid we need the cobbles or the other posters won't be very happy with me.

Have you tried using FRAPS that does a good screenshot.
Click to expand...

Actually i don't have an issue with it. As long as the screenshot is full sized and original then what difference does it make what screen it was captured on? Unless im mistaken there's no way that could be exploited. However your thread so your rules which im happy to abide by. :)
 
LtMatt said:
Actually i don't have an issue with it. As long as the screenshot is full sized and original then what difference does it make what screen it was captured on? Unless im mistaken there's no way that could be exploited. However your thread so your rules which im happy to abide by. :)
Click to expand...

Im not that botherd about it myself, what does worry me is what other people may think.

Perhaps if we could have some other peoples views as to droping the cobbles screanshot rule, if they agree we should drop it then no problem we can go with that.
 
You must log in or register to reply here.
Back
Top Bottom