FarCry Performance Revisited: ATI Strikes Back with Sh 2b

[russo121]

This is only part of the conclusion at xbitlabs

http://www.xbitlabs.com/articles/video/display/farcry20b.html

We will have more benchs with AA and Aniso active later.....

"...It is important to point out that ATI’s RADEON X800 XT and PRO graphics cards handle extreme geometry load better compared to NVIDIA’s GeForce 6800 Ultra and GT, which may mean that from this point ATI’s visual processing units have higher future-proof than NVIDIA’s latest graphics processing units do.

Additionally, ATI’s RADEON X800 XT traditionally calculates complex pixel shaders faster than everything else available today. Unfortunately, the RADEON X800 PRO still lags behind the competing GeForce 6800 GT product, whereas the RADEON 9800 XT does not seem to be a really strong rival for the GeForce 6800, at least, in FarCry...."

I'm waiting for v1.3 for sure

 

[Razor1]

was there another thread like this? hmm maybe another forum don't remember

 

[Pete]

FarCry with the latest patch 1.2 has brought two new rendering paths to support the latest graphics processors: Shader Model 2.0b for ATI’s RADEON X800, X600 and X300 series

If the X600 and X300 are PS2.b, wouldn't the 9600 also support it? Or was this a typo?

 

[trinibwoy]

Wow, whoever was saying that geometry instancing would have no effect needs to rethink things a bit. Looks like it gives a massive boost to both the R420 and NV40 in some situations.

As you see, the impact of geometry instancing is colossal: 24% performance boost in a scene with truly extreme geometry load is something terrific. It is important to note that without geometry instancing all contemporary graphics cards seem to be CPU-dependant in 1024x768 and 1280x1024 resolutions, which means that on a processor less powerful than AMD Athlon 64 3400+ there will be even higher performance advantage from geometry instancing in case of hardcore geometry load.

 

[jvd]

can't wait to see the fsaa tests . Thats all thats important . some nice 4x /16 aniso would be nice

 

[Ostsol]

A bit off topic, but: what's with the lighting difference between ATI and NVidia on the trees?

ATI
NVidia

 

[Razor1]

A bit off topic, but: what's with the lighting difference between ATI and NVidia on the trees?

ATI
NVidia

One thing that was noticed in the SM 3.0 path the lighting seems to be a bit dimmer. Now I don't know if this is the same thing in these pictures, because this is a much bigger change then what was shown before. This could also be because of LOD or a bug, looks like a bug lol.

 

[trinibwoy]

can't wait to see the fsaa tests . Thats all thats important . some nice 4x /16 aniso would be nice

I'd expect the PRO to at least catch up to the GT and the XT to put some distance on the Ultra once AA/AF is applied. Impressive speed bumps all around though. Who knows what magic 1.3 will bring

 

[jvd]

can't wait to see the fsaa tests . Thats all thats important . some nice 4x /16 aniso would be nice

I'd expect the PRO to at least catch up to the GT and the XT to put some distance on the Ultra once AA/AF is applied. Impressive speed bumps all around though. Who knows what magic 1.3 will bring

no clue hehe . But yea i suspect the same

 

[Blastman]

Something doesn’t seem right with the xbit results.

How could ixbit get big increases for the X800 on the Research level… and xbit get virtually nothing….????

 

[Heathen]

Something doesn’t seem right with the xbit results.

How do you know it's xbit getting the results wrong?

 

[ChrisRay]

Something doesn’t seem right with the xbit results.

How could ixbit get big increases for the X800 on the Research level… and xbit get virtually nothing….????

Just depends on where you are in the level.. some spots show no improvement. Some spots show massive improvement.

 

[Ylandro]

Wow, whoever was saying that geometry instancing would have no effect needs to rethink things a bit. Looks like it gives a massive boost to both the R420 and NV40 in some situations.

Yes and no.

Keep in mind, that to get those boost with geometry instancing, some major changes were made to the normal rendering!

Normally, sprites are used for far away trees and stuff. In this case, this behaviour was altered, and ALL the trees were rendered in 3D. Thus, a best-case scenario was created for instancing. And indeed, then you see a big boost.

BUT, you don't get this boost using the normal render path, when sprites are used.

So while it is obvious than instancing can boost performance, it's is equally true that it's not helping for current games, as the problem of rendering so much stuff in the distance is solved equally effective using sprites.

And comparing the screenshots of sprites and instancing, it's not a big difference, but I think I prefer the sprites image...

Could be because the 3D trees still seem to be oriented in the same direction, so it doesn't get a visual benefit over sprites. Orientating the trees in random directions, would make the trees look less alike.

 

[Ylandro]

A bit off topic, but: what's with the lighting difference between ATI and NVidia on the trees?

ATI
NVidia

Not only the trees show a difference. Look at the water edge. The 6800 show a bright line. You see it both at the beach, but also at the right, where the vertical rocks go into the water. As if the water doesn't extend fully to where it should be...

 

[Martin Eddy]

It wouldn't be PP being used would it?

 

[Veridian3]

Something doesn’t seem right with the xbit results.

How could ixbit get big increases for the X800 on the Research level… and xbit get virtually nothing….????

Could also be something to do with the fact they are using drivers which dont fully support all of the goodies in FC 1.2...

 

[Lezmaka]

A bit off topic, but: what's with the lighting difference between ATI and NVidia on the trees?

ATI
NVidia

Not only the trees show a difference. Look at the water edge. The 6800 show a bright line. You see it both at the beach, but also at the right, where the vertical rocks go into the water. As if the water doesn't extend fully to where it should be...

At least on the beach, the edge of the water is actually pretty much the same in both shots, so it's something with the beach in that area.

Another difference is that on the SM3 shot, the trees extend into the distance much farther than in the SM2 shot. I do remember them saying that using instancing for foliage allowed them to push back the view distance. I hadn't seen anyone comment on seeing the difference, but it looks like it's there on this shot.

 

[martrox]

The problem I have with the 3.0(nV) pic is that some of the trees are much darker, looking like there's no light on them....it just looks wrong.

 

[jimmyjames123]

Does anyone know why FarCry doesn't include any provision for SM 2.0a for the GeForce FX series, but includes provision for SM 2.0b?

 

[Ylandro]

No, it's not the beach itself. The area that's affeacted seems to be a blending of water and beach.

Haven't played the game, but I guess that's a part where the water is transparant? Maybe that's going wrong? Would also explain the behaviour on the steep rock wall on the right.


Seeing trees further in the distance is by game design when using instancing. Compare the ATI at 2.0b and sprites distance 100. (ATI 2.0b, sprites distance 100)