New Dx11 Benchmark StoneGiant Released

[japamd]

I took some screenshots showing tessellation levels in wireframe mode, it's amazing to know that so many polygons are being rendered in High tessellation mode !

OFF - MEDIUM - HIGH

OFF - MEDIUM - HIGH

[W1zzard]

download mirror up at tpu http://www.techpowerup.com/downloads...DX11_Demo.html

[saaya]

bitsquid tech?
fatshark?
never heard of either one of them...

1. where the FCK is futuremarks dx1 benchmark????
2. is this an ati/nvidia demo disguised as a third party benchmark?
3. why would insects attack a stone golem... and why would a stone golem care and fight them off?
4. his face looks stupid, and the spider thing is weird, besides that, very nice looking!

I took some screenshots showing tessellation levels in wireframe mode, it's amazing to know that so many polygons are being rendered in High tessellation mode !

that youd never knew were there unless you use wireframe mode

[Det0x]

This is my scores with 5870@975mhz/1300mhz + Xeon w3520 (I7 920) @ 195*21 (4095mhz).

Tessellation high on all runs.

1024*768 = 1 Run -> 61fps average
1024*768 = 2 Run -> 62fps average
1024*768 = 3 Run -> 61fps average

2048*1536 = 1 Run -> 48fps average
2048*1536 = 2 Run -> 48fps average
2048*1536 = 3 Run -> 48fps average

Same hardware settings + cata 10.3

1024*768 = 1 Run -> 62fps average
1024*768 = 2 Run -> 62fps average

1600*1200 = 1 Run -> 49fps average
1600*1200 = 2 Run -> 49fps average

Looks like driver dont matter that much. Anyway iam going back to cata 10.2
I use a samsung 1100mb (21"crt) and this guide dont work with cata 10.3

So iam stuck @ 1600*1200 with 60hz, and with 10.2 i could use 1600*1200 @ 100hz or 2048*1536 @ 85hz.

*edit*

Forgot to say the "Use extended display identification data (EDID)" in calatalyst 10.3 dont work.

[blindbox]

I took some screenshots showing tessellation levels in wireframe mode, it's amazing to know that so many polygons are being rendered in High tessellation mode !

OFF - MEDIUM - HIGH

OFF - MEDIUM - HIGH

Can you post screenshots without the wireframe? It seems like the graphics look like crap to me. Plus, what's the point if the difference is only visible in wireframe?

[tnc99]

It crash, still after installing newest dx11?

[madcho]

nice bench.

51 fps in 3 runs with q6600 @ stock win7 64bits & Ati Radeon HD 5870 cat 10.3. 8 Go DRR2 800. ( all settings by default. )

( running with some apps in same time, i've don't optimized ).

Graphics are nice.

[kaktus1907]

Ridiculous what DOF does to fps, especially with no noticeable IQ gain..
DOF kills Metro as well..

[XSAlliN]

1. where the FCK is futuremarks dx1 benchmark????

They lost their touch. Have you seen 3D Mark Vantage? Those graphics look like they where made to look impressive for year 2003. Getting low FPS with High-End systems on those ty graphics, was obviously bad optimizations.

[japamd]

bitsquid tech?
fatshark?
never heard of either one of them...

1. where the FCK is futuremarks dx1 benchmark????
2. is this an ati/nvidia demo disguised as a third party benchmark?
3. why would insects attack a stone golem... and why would a stone golem care and fight them off?
4. his face looks stupid, and the spider thing is weird, besides that, very nice looking!

Wow, bad day, saaya?

that youd never knew were there unless you use wireframe mode

This one has very perceptive differences in normal rendering mode too, I think. I'll do what blindbox is asking below so we can compare those tessellation off/medium/high modes.

But I'm guessing they're choosing the wrong ambients to showcase tessellation, they're almost dark scenes (AvP, Metro 2033, this techdemo) or fast paced environments like in Dirt 2 where those differences are little noticeable. A game like Oblivion would show better results IMO.

Can you post screenshots without the wireframe? It seems like the graphics look like crap to me. Plus, what's the point if the difference is only visible in wireframe?

I'll do that later, besides graphics being not so good, tessellation adds a very good level of detail to them and that's noticeable in normal rendering mode too.

[Manicdan]

what does depth of field change? seems to be a big perf hit, but are the effects worth it?

[Xello]

Nice demo, but i think the wait for a truly impressive implementation of tessellation continues.

[DosDuoNo]

Average FPS (DOF off): 75
Average FPS (DOF on): 42

This was with a GTX480 @ 800/1600/3700 (197.44 drivers) and Core i7 920 @ 3.6GHz and 6GB DDR3 1600 RAM

1920x1200, 16:10, full screen, tessellation high

Average FPS (DOF off): 69
Average FPS (DOF on): 36

This was with a GTX470 @ 608/1215/3348 (197.44 drivers) and Phenom II @ 3.5 gigand 4GB DDR2 1066 RAM

1680x1050, 16:10, full screen, tessellation high

i liked your format so thought i would just quote it for some comparison between 470/480 and AMD/Intel

[japamd]

I suppose DOF is on by default? How do you guys are turning DOF on?

Ooops... nevermind, found in readme.txt :P

[japamd]

Tessellation levels compared

OFF - MEDIUM - HIGH

OFF - MEDIUM - HIGH

OFF - MEDIUM - HIGH

OFF - MEDIUM - HIGH

OFF - MEDIUM - HIGH

[Final8ty]

They lost their touch. Have you seen 3D Mark Vantage? Those graphics look like they where made to look impressive for year 2003. Getting low FPS with High-End systems on those ty graphics, was obviously bad optimizations.

Indeed i think the gfx in 3DMark 06 look much better.

[justin.kerr]

rig in sig 1920x1200 high settings DOF on

[Chickenfeed]

At 1680x1050 with high tessellation, 3 run average, system in signature :

29fps ( diffusional depth of field on )

42fps ( diffusional depth of field off )

As far as I can tell though turning off diffusional dof ( F3 key ) doesn't completely turn off dof however. In the prerelease version, depth of field worked on ATI hardware but not diffusional depth of field.

[justin.kerr]

your #'s seem low?

[Chickenfeed]

Its inline with what I got using the prerelease (normal dof) one so no idea... Peoples numbers look very inconsistent though, they varry way too much system to system so who knows.

Like for example DosDuoNo did his at 1680x1050 on his 470 and got 36 which I'd believe given Fermis prowess at tessellation.

Its possible it doesn't like the 10.4 drivers but no idea, this system still gets 10k on vantage extreme so i know its working properly

[justin.kerr]

did you run it in full screen?
I am running the 10.4a's

[Chickenfeed]

Yes full screen. I tried windowed mode for the hell of it and it made no difference.

[justin.kerr]

the FPS in game, and what FRAPS say are way different, I just checked that..

[Chickenfeed]

I did 3 runs with fraps running along side in bench mode and 1 not in bench mode ( so I could still see the counter ) and both fraps as well as the ingame counter both showed the same numbers for me. When it comes to multi gpu setups though in game frame rates and even fraps can be misleading.

Could you try running with Catalyst AI off and at 1680x1050 perhaps? This should provide a more relavent comparison.

[{SAS}TB]

Hi

I ran at 1680x150 - DOF off and High Tesselation and got



Seemed a tad odd and ran the BM taking pics at random times of the FRAPS counter:



Something looks off to me ?

Nice looking BM though

same settings DOF and Tesselation off