UNIGINE-Valley BENCHMARK scores

[Soulmirror]

Trying to do my best but still i can't beat you D007

Asus Rampage V Extreme 3.1
i7 5930k @ 4.5
Nzxt Kraken X61
32GB Ripjaw V DDR4-3000
2x Zotac 980Ti Amp! Extreme 1450/1810
Samsung 850 512GB
2x Samsung 850 1TB Raid 0
Corsair HX1000i

Maybe i should go higher on the memory side, let's try again tomorrow.

 

[catnipkiller]

i7 4790k stock 2 gtx670s stock
Score: 4229

 

[jboydgolfer]

 

[RealNeil]

View attachment 70059



Trying to do my best but still i can't beat you D007

Asus Rampage V Extreme 3.1
i7 5930k @ 4.5
Nzxt Kraken X61
32GB Ripjaw V DDR4-3000
2x Zotac 980Ti Amp! Extreme 1450/1810
Samsung 850 512GB
2x Samsung 850 1TB Raid 0
Corsair HX1000i

Maybe i should go higher on the memory side, let's try again tomorrow.

You're both getting great scores!

 

[Fouquin]

Radeon R9 Nano

 

[jboydgolfer]

Radeon R9 Nano

does that pentium hold you back when gaming with that fury?

 

[Fouquin]

does that pentium hold you back when gaming with that fury?

Not that I've seen. Fallout 4 and Armored Warfare at least don't take a penalty.

I'm about 2 weeks away from getting another 4690K though, so I guess I'll find out if it really does hold it back.

 

[Gregory Hartley]

i7 4790K - OC 4.6 Ghz
Gigabyte GTX 970 G1
GPU Clock - 1318/1469
Memory Clock - 1928 MHz (7,712 Effectively )

 

[PP Mguire]

Core i7 3960x @ 4.9GHz | 2x GTX Titan X 1468\1920

 

[RealNeil]

A lofty score indeed!

 

[PP Mguire]

It's not even 100%. I couldn't get my cards to 99% each like Heaven and I gave up trying last night. AFR 1 and 2 both would max the cards then in certain scenes drop back to about 50. If anybody has any ideas LMK. AFR 2 sometimes crashed the program.

 

[DarthBaggins]

I need to run valley now lol

 

[jboydgolfer]

jesus 2 titans, i couldnt afford half of one Mucho envyyyyyy.. My PC is almost as old as my youngest kid

 

[DarthBaggins]

I'm happy with just my lil 970 (especially since I got it thanks to MicroCenter's awesome extended warranty plan I had on my Asus HD7870)


Ran my Valley bench finally:
5930k @ 4.4 Asus Strix GTX970 @ 1501/3557

 

[SpikeHob]

 

[Ferrum Master]

The table here is such mess, needs better sorting. SLI and single in different tables. Can't really understand where to aim.

@PP Mguire post your single Titan X Score... just for science

For start CPU 4.7GHz and GPU/MEM 1510/2005

I want to see your scaling score with single... and then put again to 5GHz... actually the CPU bumps up not very much, the GPU core... and actually RAM frequency does lately some mojo...

EDIT: I just realized your double titan score is.... something is wrong...

 

[PP Mguire]

The table here is such mess, needs better sorting. SLI and single in different tables. Can't really understand where to aim.

@PP Mguire post your single Titan X Score... just for science

For start CPU 4.7GHz and GPU/MEM 1510/2005

I want to see your scaling score with single... and then put again to 5GHz... actually the CPU bumps up not very much, the GPU core... and actually RAM frequency does lately some mojo...

EDIT: I just realized your double titan score is.... something is wrong...

View attachment 70722

Yes sir, both GPUs were only running 48% with that run and idk why. No matter what I did it would not do a full run maxed like Heaven.

 

[Ferrum Master]

Yes sir, both GPUs were only running 48% with that run and idk why. No matter what I did it would not do a full run maxed like Heaven.

For the lulz did try DX9 or OGL? Seconds... old exe renaming trick? I guess no use... there is something broken. Or... it is just the limitation or the DX11?

 

[PP Mguire]

For the lulz did try DX9 or OGL? Seconds... old exe renaming trick? I guess no use... there is something broken. Or... it is just the limitation or the DX11?

It's just this program as Heaven runs fine obviously. I'm not real sure what the deal is but it's not a game so not a deal breaker. If we're doing single card anyways no reason to really try hard to get it going right.

 

[Ferrum Master]

It's just this program as Heaven runs fine obviously. I'm not real sure what the deal is but it's not a game so not a deal breaker. If we're doing single card anyways no reason to really try hard to get it going right.

Well actually heaven still runs like shit considering what we are throwing at it. I guess this is something bigger. It goes deep into task queues for specific jobs. How they are stacked up and serialized, I guess some parts just don't or can't be run in parallel, thus waiting and some budget time, the next frame maybe is discarded. It would be fun to see some uber high clocked Anniversary Pentium here again... see if it really isn't tied to poor multitasking and relies to brute force.

 

[PP Mguire]

Fun fact, one of my Titans can run over 1500 core with 2000 mem. Interesting. My second Titan isn't matching vcore with it being .01 lower than the first. My score is like 175 under yours btw @ 4.9 1501/2000.

 

[Aquinus]

@PP Mguire, have you tried re-running it with your CPU at stock? Is it possible that you're encountering a CPU bottleneck? An earlier run someone else did with two 980 Tis isn't too much higher than your score (600 points?) and it is making me wonder if something other than your GPUs is a bottleneck. If your score in SLI is lower at stock speeds, then that's a tell tale sign that you've hit a limitation with your CPU (probably a single thread getting maxed out.) Many applications become CPU bound at really high frame rates.

Also, you have to admit, that's a lot of GPU horse power to drive 1080p. I suspect if you ran the benchmark at 4k it would eat up more GPU power which would also indicate a CPU bottleneck.

 

[Ferrum Master]

Fun fact, one of my Titans can run over 1500 core with 2000 mem. Interesting. My second Titan isn't matching vcore with it being .01 lower than the first. My score is like 175 under yours btw @ 4.9 1501/2000.

I am running custom edited bios... mine does not fiddle... it is always on max boost speed if I put power limits in afterburner to max. And my RAM speeds are higher than yours... that's the only difference.

 

[PP Mguire]

@PP Mguire, have you tried re-running it with your CPU at stock? Is it possible that you're encountering a CPU bottleneck? An earlier run someone else did with two 980 Tis isn't too much higher than your score (600 points?) and it is making me wonder if something other than your GPUs is a bottleneck. If your score in SLI is lower at stock speeds, then that's a tell tale sign that you've hit a limitation with your CPU (probably a single thread getting maxed out.) Many applications become CPU bound at really high frame rates.

Also, you have to admit, that's a lot of GPU horse power to drive 1080p. I suspect if you ran the benchmark at 4k it would eat up more GPU power which would also indicate a CPU bottleneck.

Both cards run around 48% even in 4k. The person with dual 980ti's are probably having the same issue I am. The cards still run 48% with the CPU stock. When I change to AFR1 or 2 different scenes will max them but will cause the program to crash even at stock meaning I'm suspecting just a SLI profile issue of some sort. Since my install is old I gather something I've done over the last 6 months isn't playing nice.

I am running custom edited bios... mine does not fiddle... it is always on max boost speed if I put power limits in afterburner to max. And my RAM speeds are higher than yours... that's the only difference.

I'm running custom HC bios on my cards. I kept my RAM stock because in game it doesn't make a difference in FPS. The last run I did had my RAM at 2000 and it only raise like 20ish points over 1950. My stock boost speed is 1391 so I only have to raise 110 or so on core to get over 1500, but my second card will not run that high due to the minor difference in voltage. They are both 100% stable with the HC bios on "stock" though.

 

[broken pixel]

I will bench my 2x 980ti's when I get back to my puter.