Unigene Valley Extreme HD

[alienware]

Well after seeing Tom playing with his Titan I figured it was time to finally run this and see what score I got.

Specs are in sig and should be about right, though I have a 1200w psu so excuse the 1250 typo.

I set it to Extreme HD and ran it though three times, saving my best score.

What's very odd is that I am not overclocking my cards yet in the top right corner they are both apparently clocked to 1254mhz. Very strange !

So yeah, basically run the benchmark and post your results making sure to let us know the spec so we can compare things properly

 

[SeekaX]

Well after seeing Tom playing with his Titan I figured it was time to finally run this and see what score I got.

Specs are in sig and should be about right, though I have a 1200w psu so excuse the 1250 typo.

I set it to Extreme HD and ran it though three times, saving my best score.

What's very odd is that I am not overclocking my cards yet in the top right corner they are both apparently clocked to 1254mhz. Very strange !

So yeah, basically run the benchmark and post your results making sure to let us know the spec so we can compare things properly

that's a problem with the program apparently. my 670ftw is at 1150mhz with kelper boost and unigine says i'm at 1250 as well.

 

[alienware]

Oh OK cool I was beginning to think the previous owner had tampered with the bios on the second card (was a return).

So come on then, get the benchmark running and show us the score

 

[SeekaX]

Oh OK cool I was beginning to think the previous owner had tampered with the bios on the second card (was a return).

So come on then, get the benchmark running and show us the score

already have it got all the benchmarks and the pictures lined up, just waiting for some time i can write the review.

 

[alienware]

Cool beans I look forward to that !

 

[Scoob]

Not near my gamer now, but from Memory I get about the following from my Sig Rig:

FPS: 81
Score: 3400
Min FPS: ?
Max FPS: 150

Just from memory, I will check later.

One thing to add. I run at 1200p natively, but popped the settings to 1080 in Valley to compare and scored less than my 1200p score - go figure.

Scoob.

 

[alienware]

Seems about spot on for a pair of 680s. Nice to see some one else finally run it lol.

 

[Scoob]

Just did a run with the new 314.22 WHQL drivers:

FPS: 82.8
Score: 3462
Min FPS: 35.5
Max FPS: 154.3

Preset shows as "Custom" as I run at my desktop res of 1920x1200, Quality (underneath Preset) shows "Ultra" I notice your screenshot does not.

I've hit 1200mhz (boosted) on stock volts on my EVGA 680's, and 7ghz on the vRam. I have considered doing the BIOS tweak/update to unlock the voltages but, really, I'm yet to see the need...though I do feel the desire lol.

Edit: I also did a quick run in the main Heaven 4.0 Benchmark, again at 1920x1200:

FPS: 70.7
Score: 1781
Min FPS: 28.7
Max FPS: 154.2

Of course this is with 8xAA, Ultra Quality and Extreme Tessellation. The newer drivers seem to have made Heaven a little smoother to the eye, though overall scores are near identical.

Scoob.

 

[alienware]

Extreme HD is a preset in Valley and shouldn't be changed.

 

[Scoob]

Extreme HD seems to lock you to 1080p though, I run a 1200p monitor...those extra 120 pixels make ALL the difference lol.

Scoob.

 

[alienware]

Extreme HD seems to lock you to 1080p though, I run a 1200p monitor...those extra 120 pixels make ALL the difference lol.

Scoob.

Oh sorry mate I see what you mean now.. Yeah when I originally ran it it set the resolution to native which is what my TV takes and it's something oddball like 1790x1020 due to overscan or underscan? something like that..

Running at 1080p locked using Extreme HD lost me about 350 points..

The reason I said to use Extreme HD was so that we could compare the results accurately to see how certain cards in certain situs perform

 

[Scoob]

Happy to do a run at 1080p, however as mentioned earlier, I noticed that my score is slightly worse than 1200p...still, I've updated the drivers since so maybe that's changed.

To be fair, I'd expect minimal differences between 670's and 680's in most things.

I will do a 1080p run of valley & edit this post...

Edit: Ok, just did a run at 1080p with the "Extreme HD" preset.

FPS: 89.0
Score: 3722
Min FPS: 37.2
Max FPS: 164.0

Looks like the driver update has fixed the ealier anomaly. Nice.

Scoob.

 

[alienware]

Which driver did you change to?

 

[Scoob]

Which driver did you change to?

I was on one of the Betas, but updated to 314.22 WHQL the other day. I cannot remember exactly which driver I was on when I notice the 1080/1200p anomaly though.

Scoob.

 

[FTLN]

Just thought I'd chip in with my 7950 results

3770k @ 4.7
7950 @ 1100 / 1575 / 1200mv
4gb cheapo ram

 

[alienware]

Good result that. Sadly I can't overclock my cards as one of them doesn't like it.

 

[Scoob]

Good result that. Sadly I can't overclock my cards as one of them doesn't like it.

That's a shame. Best I managed was 1230mhz on the core of each of my 680's - I run them at 1200 though so they're not at their absolute limit. I was more impressed by the 7ghz vRam! Not tried to push that further.

Still, even at stock a pair of 670's is pretty potent!

Scoob.

 

[alienware]

I had 1250 out of the first card I bought easily. The second one was used on Amazon and was probably returned as it doesn't overclock very well at all. We're talking 60 mhz if you're lucky.

Normally I would return it but there's not much point in SLI on air. It just adds to the temps which are pretty wild as it is. Without overclocking they remain under 72c.

 

[Scoob]

I get how frustrating that might be though. My original Palit GTX 570 (reference) was the best of the 570's I had. Lower vCore, higher clocks as well as cooler and quieter. The 2nd hand Inno3d I bought couldn't really keep up, so I ran both at 800 or there abouts when the Palit could do 930 happily. Ironically it was the Palit that died when I had a compression fitting fail.

Normally I'd go for the 470/570/670 rather than the '80 card due to the value for money aspect, they usually offer 90% of the performance for 2/3rds of the price. This time round however the 680 was only £40 more expensive than the 680 (for the EVGA's I wanted) so I treated myself to the more expensive card.

Do you think you'll water cool? Possibly your weaker 670 will be happier running a little cooler?

Edit: Here's an interesting result. I just ran Valley at Extreme HD on my old EVGA GTX 570 @ 850. It's now sat in my Media PC (Q6700 @ Stock) and my scores were just under half that of my two 680's. In lots of tests I saw prior to getting my 680's ONE 670 usually scored the same as my two 570's. Amazed how well the old single 570 is scoring really.

Scoob.

 

[alienware]

I'm not planning on water no. The case would more than likely accommodate it if I wanted to (all be it messy) but there's not much point. I have really good custom cooled cards and the top one gets to around 72c when being pushed really hard.

I'm not a fan of water cooling because I have terrible OCD and I'd be forever worried it was going to leak :lol:

Not made easier by having a AIO cooler leak on me and ruin a GPU and board.