DX 11 benchmark for ATI guys

I couldn't get it to run in DX11 either.

I was getting over 100FPS in the benchmark (usually between 80 and 100) and it was pretty choppy

 

DX11 worked here.

 

Cant wait to get 4 cards overclocked and dry ice on the proc for this bench.. still waiting for rma. this week I hope

 

That's possible but it's not called a benchmark for nothing.

A bench mark is a gauge to check how one's hardware performs. OKay it's a big maybe, the developers of this bench did not code it properly, well they are not doing anyone a favor since it's the first dx11 bench.


Somehow from the look of things the very first games that use dx11 are probably going to act like this bench which is get lower frames when tess and dx 11 are enabled.

 

Yes totally agree with you on that the benchmark does not impress I mean my bro came in add a look at it he thought it looked like xbox 1 graphics and I am like it uses tess he said I don’t care I have seen a lot better 2 years ago.

When someone says that after the graphics on display are using state of the art direct x 11 you know something is not right I reckon dirt 2 with DX11 will be fine.

Are you using windows 7 bit strange you can’t get DX11 part to work I am using vista just had the final update yesterday for DX11 to bring it up to windows 7 standard plus are using the latest drivers off the AMD site.

Do you want to here something great I emailed AMD asking why I can’t get there DX11 demos to install on vista, add the reply today they say because windows 7 is the only OS to support DX11 vista is only DX10 I mean they made all the videos saying vista and windows 7 for DX11 gaming and all that, I have emailed them back showing how stupid this guy is who emailed me.

 

why does this benchmark only use 1 cpu core ? , is this a feature of dx11 ? . seems odd if you ask me

 

must be to stress the gpu more..

 

maybe but i would expect the 5870 to get a higher score than it does tbh , seems like the fx5800 ultra all over again . that performed bad in dx9 but great in dx8 etc , surely a card (5870) with this kind of power would perform better

 

That's also to be said about the 8800 GTX, and the cheaper 8800 640 MB and 8800 320 MB GTS. Could they run the first ever DX10 3d Game?
I mean, it was a card "with that kind of power".

I do not think that will happen in DX11 because of performance tweaks. Also, the tesslation in this engine is the first to go public as a DX11 benchmark. I shouldn't worry.

 

How's that? Nothing funny in it and they are running it using DX11 API. Heck, even my laptops Intel GMA 4500MHD can run it in DX11. The point in DX11 is that it's comptaible with DX10 cards and it actually even adds features to DX10 cards like compute shader 4.0.

 

http://www.pcgameshardware.com/aid,...PGs-and-strategy-titles-might-look-like/News/

Some more details on DirectX 11 via a updated build of the Trinigy engine.

EDIT: Seems I mixed the various names together a bit, Unigine, Trinigy and what else.

 

Just can't get it to run ???

5870, Catalyst 9.10, newest DirectX and Win 7 Home Premium

What gives ???

 

Here's mine
OC 1000/1250

 

HardOCP just did a review of the benchmark.

Conclusions:
1: Tesseelation does more overall work, and sustains at least a 20% FPS penalty.
2: The DX11 API is around 1% faster overall (never more then 3-4 frames) compared to DX10

Can someone on Vista do the following [assuming a DX10 card]:
1: Run without the DX11 API being installed
2: Run after installing the DX11 API

I want to see if the supposed multithreading in DX11 makes any effect on DX10 hardware.

 

This is a pathetic benchmark at best...I ran DX9, DX10 and DX11....my framerates never dropped below 40FPS....they RARELY dropped below 50FPS, yet it decided my "average FPS" was 46.9 in DX11, 47.8FPS in DX10, and 49.9 in DX9. In most tests, I was staying between 58FPS and 70FPS....I'd dip into the 40's for 2-3sec per "stage". In the DX9 benchmark, my lowest was 52FPS...how was my "average FPS" below my lowest FPS from the entire benchmark?????

Funny to see people running ATI cards making such statements seeing as how DX10 contains most of the DX11 instructions, and the remaining instructions (with the exception of Tessellation) can be emulated by CPU. Guess you should have read the FAQ sheet on DX11. We may take a small performance hit due to the emulation, but unless trying to use Tessellation, we still get DX11.

Because it wasn't written to fully use the DX11 API....which was designed for multi-core systems. They have intentionally limited it to 1 cpu core.

 

but why bottleneck the gfx ? , this will not give true performance of the 5*** series ? . i always thought a benchmark was to test the gfx etc to see what they could do

 

why would you run a 5800 series card on that res?

 

bad performance

My OC'ed XFX Black edition 5970 is sad for performance if MSAA is on period it lags like a little baby.

The GTX 480 I benched in my system took on the Heaven Benchmark like it was nothing maxed with MSAA on 8X. It took my card OC'ed at 900/1200 to just get 2k more points over the GTX in 3DMarks. I have a beast of a system as well.
I got with the ati card a range from 32-38 average FPS in heaven bench min 11 max 140 the dragon seem crushed my card but was flawless with the gtx

System Specs:
Cat 10.6
* Intel i7-930 @ 4.30 Ghz
* Asus P6X58D Premium BIOS 0703
* 6GB Kingston HyperX T1 Series KHX2000C9AD3T1FK3/6GX @1683Mhz
* Corsair H50 Cooler Push/Pull 133CFM
* Antec 1200 w/ 6 133CFM 120mm fans and top 200mm stock

You can see the GTX bench on my blog still uploading the ati bench.
http://www.therandomtechblog.com/

ati at stock settings 3dmarks06 pro


gtx 480 at stock


We did about the same in Vantage and Heaven

Now in Metro 2033 maxed and in DX11 I get about 60FPS where his did 27-35FPS with MSAA on or off I can't run the game period with multisampling on.

I want to know why this beast of a card and hole in my pocket doesn't perform well at all also I am using an Arctic Cooling Accelero Extreme for my ati

 

Heaven benchmark is tuned for Nvidia cards, and guess what? Metro 2033 is also tuned for those cards. When you look for benchmarks on those two cards you'll see that you picked the two titles that the GTX480 is faster then the HD5970.

Edit: Why did you take 3dm06 as a reference? It's heavily cpu dependant. Take actual game benchmarks or even Vantage as a tool to look at the difference between those two cards.