***Official Unigine Heaven 4.0 benchmark scores! Thread***

[futurefields]

---------------------------
Sapphire 7950 @ 1100/1500

 

[Durvelle27]

we have one bud

http://forums.anandtech.com/showthread.php?t=2301583&highlight=heaven

 

[futurefields]

That one has people running in different settings and stuff, like for example you posted your score with Normal tessellation but you are supposed to use Extreme tessellation for benchmarking. It would be nice if there is a thread where everyone simply posts their results using the same benchmark settings. And then as they are added I can add them to the bottom of the 1st post, so we have a list.

 

[Eureka]

Heaven already comes with a Extreme preset. Why not just use taht?

 

[futurefields]

Because it uses a "windowed" view which seems to give people wildly different results. The fullscreen test is much more consistent.

 

[Durvelle27]

HD 7870 1200/1450

 

[Fastx]

Benchmark setup:

1920x1080, Fullscreen, 8x Anti-Aliasing
Ultra Quality
Extreme Tessellation

Sapphire 3L 7950
1100/1400
1.181v
BP +15
Temp 64c


Notes
I only wanted to try 1100/1400 I don't have VRAM sensors to monitor Vram temp on this card so decided to only take it to 1100/1400 only. This is as high as I had to plan to go (1100) before I bought the card, just interested in a mild oc like the above for games only then back to stock settings 24/7 when not gaming. Don't want to push the card I am really not into max OC.

 

[futurefields]

Fastx are you sure it's 1.8 volts? That seems high for a 7950?

 

[Fastx]

Fastx are you sure it's 1.8 volts? That seems high for a 7950?

No its not I reedited it.

 

[Face2Face]

Might as well post mine here too.

When I get more time I will try to tweak it and see if I can get a better score. Heaven 4.0 is a lot tougher to run then Valley..

 

[imaheadcase]

Because it uses a "windowed" view which seems to give people wildly different results. The fullscreen test is much more consistent.

Most don't use windowed view though..

 

[BallaTheFeared]

A bid for a thread with sane settings.

 

[futurefields]

Balla your results you posted are using different settings, the point is everyone uses the same settings, so one browsing the thread can quick and easy compare.

Please retest using the settings in the OP and replace the old results image with new one, thanks.

 

[futurefields]

Might as well post mine here too.

When I get more time I will try to tweak it and see if I can get a better score. Heaven 4.0 is a lot tougher to run then Valley..

Thanks, are these results for the same OC settings listed in your sig? Im going to start compiling the list when i get home from work, will look like -

Score - card/settings - username

And i will post the results in descending order.

 

[BallaTheFeared]

/\

 

[futurefields]

This is how benchmark work for everybody Balla

 

[BallaTheFeared]

I was hoping for a non ROP/Tess Mark version /\

 

[futurefields]

This will only work if we all use the same settings.

 

[Face2Face]

Thanks, are these results for the same OC settings listed in your sig? Im going to start compiling the list when i get home from work, will look like -

Score - card/settings - username

And i will post the results in descending order.

Makes sense. I forget right now, I think it was 1300/1800 - I will check later tonight. I assume you will post results in your 1st post?

 

[futurefields]

Yes I will edit the OP and add them all as they arrive.

 

[cmdrdredd]

ignore the part where it says 1x next to my GPU. I am indeed running SLI. Both cards loaded to ~99%

Single Card

 

[Face2Face]

^ Hey Dredd, what clocks were you running in SLI and single card? Nice SLI score BTW!

 

[cmdrdredd]

^ Hey Dredd, what clocks were you running in SLI and single card? Nice SLI score BTW!

1215Mhz Boost, 6732Mhz Memory

Both in SLI and Single I used the same clocks. That's the max my first card can handle without some issues in some game. So I adjusted my second card to boost to the same level.

Card one needs +60 to boost to 1215 and card two needs only +33.

 

[futurefields]

Thanks guys, I updated the OP with the score results so far!

 

[Face2Face]

Just wanted to confirm my clocks were at 1300/1800.

Come on people let's get benchmarking!!!