New Unigine "Valley" Benchmark

[auxy]

Eh, they were somewhat acceptable from a hardware point of view and price/performance point of view back at their day... Of course, the "performance" part didn't apply to some games because of half-assed (which was typical for this company) driver optimizations and the long, long time it usually took for those optimizations to materialize - I remember having multiple issues with BF:BC2 and my HD6950 card long time ago, where some driver release was causing constant crashes of this game after about an hour (or slightly less) of playing it, and also something about slow level loading times due to specific shader compiling in DX10 mode (switching to DX9 made maps load faster)... Anyways, was my last ATI/AMD card and I'm glad I removed it from my PC :wink:

My brother has crossfire 6970s in his current PC (well, BIOS-modded 6950s), and he constantly complains about them -- and then he'll turn around in the next breath and tell me how awesome they run whatever game. I've seen them perform adequately, but I've also seen the issues he's had with them.

He won't give me the pleasure of displaying open envy for my machine, but he's already parting out a Haswell machine with a pair of 7970GE. Heeheehee!

[Jason181]

Ultra, 8x AA, 1680x1050, DX11.

Min: 24.8
Max: 126.2
Avg: 64.8

[cwj717]

On my 7970@1200/1575:

1920x1080 8xAA:
FPS: 50.9
Score: 2129
Min FPS: 23.6
Max FPS: 97.1

http://www.filedropper.com/uniginevalleybenchmark10201303170016html

[cynan]

Last one. For real this time.

[ClickClick5]

I should have the list posted tomorrow. If you want to be in it, post your results.

[auxy]

I should have the list posted tomorrow. If you want to be in it, post your results.

You can pull my results from the thread, huh? (ﾉ)´∀｀(ヾ)

Last one. For real this time. [image]

Very nice! Within spitting distance of my results, hehe.

[JustAnEngineer]

Here's what a $95 Radeon HD7770 can do (at stock speed):

[ClickClick5]

Tada! Results! Only those who posted at Extreme HD settings and so forth were added. I did the best I could with the info posted here. It is just a simple Excel 2010 spreadsheet. Feel free to post updated info and I'll update the sheet.

http://linserv.biz.tm/dat/TR_Valley_Bench_Results.xlsx

EDIT: Save link as...

[auxy]

Here's what a $95 Radeon HD7770 can do (at stock speed): [image]

Hey JAE, could I ask you to run that for me again on 1080p noAA? 「(°ヘ°)

I've seen a few places -- most prominently Tom's -- list the 7770 as "similar" in performance to a GTX460, but your Extreme HD result is about half of what I got on my 460; I wonder if they aren't similar in performance but the 7770 is held back by its gimpy 128-bit memory at 8x AA?

Tada! Results! Only those who posted at Extreme HD settings and so forth were added. I did the best I could with the info posted here. It is just a simple Excel 2010 spreadsheet. Feel free to post updated info and I'll update the sheet.

http://linserv.biz.tm/dat/TR_Valley_Bench_Results.xlsx

EDIT: Save link as...

Unfortunately, we're required to use IE8 with some fairly strict content filtering here at work, so I can't download the spreadsheet. Clicking it brings up a nanny page, and right-clicking to "Save target as..." just saves an HTM file which is just a link to the spreadsheet (which, when opened, gives a nanny page.) I'll look at it when I get home. (^∇^)

[JustAnEngineer]

FWIW, the real-world performance of a hot-clocked GeForce GTX460 is very similar to a Radeon HD5850. I confirmed this equality when I owned both of those cards. It's possible that this Unigine benchmark, like their last effort, is a TWIMTBP tool that favors one brand over another. How did the 3DMark comparisons look?

[auxy]

FWIW, the real-world performance of a hot-clocked GeForce GTX460 is very similar to a Radeon HD5850. I confirmed this equality when I owned both of those cards. It's possible that this Unigine benchmark, like their last effort, is a TWIMTBP tool that favors one brand over another. How did the 3DMark comparisons look?

Dunno, I've never run 3DMark. :lol:

Given cynan's excellent Valley results tho, I don't think this tool is particularly biased. ┐(￣ヮ￣)┌

[cynan]

Given cynan's excellent Valley results tho, I don't think this tool is particularly biased. ┐(￣ヮ￣)┌

Well, my 7970 is under water. I'm sure if you water cooled your Titan and overclocked it accordingly, it would put things into perspective nicely - ie, reestablish a strong 30%+ lead over the HD 7970. Plus, runs where I was scoring ~55+ avg FPS were at settings I'd not run the card at everyday, and produced the odd artifact (but never became unstable). At the higher settings I'd actually game at (ie ~1250 Mhz core/1600 Mhz mem), I don't score more than 52 avg FPS. But then where is the fun in benching conservatively? This is an e-peen fest after all, right?

The reason I reposted with higher scores recently is due to the new Catalyst 13.3 beta 2 drivers. It seemed to make a notable improvement in Valley. I also tried disabling hypertheading and enabling PCIe 3.0 from recommendations on this and other forums. Hyperthreading didn't seem to make any difference. Turning on PCIe 3.0 perhaps made a slight improvement, but no more than a few fractions of an (average) FPS (but this could have been attributable to natural benchmark variance).

I don't know what's up with that low GTX 680 score in the above spreadsheet... Seems like you come across the odd out of line result regardless of whether the card is AMD or Nvidia.

[auxy]

Well, my 7970 is under water. I'm sure if you water cooled your Titan and overclocked it accordingly, it would put things into perspective nicely - ie, reestablish a strong 30%+ lead over the HD 7970.

Perhaps! My point, though, was more that your scores are very good regardless of comparisons to other cards, so I don't think the benchmark is especially biased. I actually would trust Unigine much more than 3DMark -- or anything from Futuremark -- to be unbiased. ┐(‘～`；)┌

[cynan]

I actually would trust Unigine... ...to be unbiased.

Agreed. Performance seems to scale more or less evenly across AMD's and Nvidia's more recent GPUs. If anything the HD 7900 series may have a slight edge over the GTX 67/80 cards at the Extreme HD settings due to 8x AA and the Radeons' higher memory bandwidth. But, overall, it seems pretty even.

[ClickClick5]

I actually would trust Unigine... ...to be unbiased.

Agreed. Performance seems to scale more or less evenly across AMD's and Nvidia's more recent GPUs. If anything the HD 7900 series may have a slight edge over the GTX 67/80 cards at the Extreme HD settings due to 8x AA and the Radeons' higher memory bandwidth. But, overall, it seems pretty even.

Hence why I put the excel sheet together lol! I'm not seeing massive distances between cards like in 3DMark.

[auxy]

Well, I modded my BIOS and did a few other things, and here's my new benchmark if anyone is curious:


Hehe.