Anand has Farcry 1.2 patch and tests SM3.0

[termikk]

it smells...
graphic cards business is a dirty business
what bull, a person should at least be able to plug in, install game and play in peace, but no!
shader this shader that, optimisation this, optimisation that, cheat this cheat that, corruption, patch, controlpanel, glitches...
game developers are whoring themselves for allowing this (some sites...)

Think of the normal player that just wants to play a little, in peace, no problems, no hassle.

 

[kipper67]

Which sites Chal?

http://www.xbitlabs.com/articles/video/display/farcry30.html

The firingsquad results aren't that far off, either, but you just have to remember that they used their own timedemos, and we know that the NV40's advantage comes from specific scenarios, so performance should obviously depend on which timedemo you choose. So I say their results are also similar.

I got the impression Firingsquad used one custom demo of their own, the rest, were the 4 supplied by Crytek/Nvidia along with the patch.

With all of this in hand, plus four new demos that come with the new 1.2 patch, we were eager to test out the improvements, but first let’s get started with our own custom demo, which is based on the map mp_monkeybay (which happens to have its fair share of grass in it).

From the foot of the first page.

Mark

 

[kmolazz]

Have you ppl read the xbitlabs article, very informative. Much more information than anands as they compare nvidias and ati cards in v1.1, v1.2 and with sm3.

X800 cards performance drops a little with the 1.2 patch and the boost nvidia cards get isn't only from sm3 but also some other improvements in the patch.


http://www.xbitlabs.com/articles/video/display/farcry30.html

 

[jvd]

I was allways under the impression that if a patch is a performance patch then it should increase performance for everyone . I wont be downloading this patch. no need for it and it will only slow me down

 

[John Reynolds]

Just double-checked with the 61.45s and CP enabled AA does not give AA in Far Cry 1.1.

 

[KimB]

Just double-checked with the 61.45s and CP enabled AA does not give AA in Far Cry 1.1.

For some reason, you have to set it in the application profile for Far Cry for it to work.

 

[ChrisRay]

Just double-checked with the 61.45s and CP enabled AA does not give AA in Far Cry 1.1.

For some reason, you have to set it in the application profile for Far Cry for it to work.

Actually it doesnt matter. I have set it to No AA/NoAF. Or 8xAA/16xAF .

My Far Cry will only enable AA if its set in game menus or that little config menu outside of game.

Setting it to App Preference didnt make a difference.

 

[2senile]

http://www.pcper.com/article.php?aid=55

You might want to check out the PC Perspective take on PS 3.0 & the 1.2 patch. Lots of pretty graphs.

I have to say that, as long as there is no detrimental effect on other cards, 6800 users should be thankful to get a performance increase from the "SM3" path. Doesn't seem to be an overwhelming incentive for me to upgrade from my 9800 Pro tho.

Also, I was under the impression that it was VS 3.0 that was the big deal in SM3 so, as a non-techie person, I don't understand why there isn't anything utilizing it. Why did they put the effort into PS 3.0?

 

[KimB]

Actually it doesnt matter. I have set it to No AA/NoAF. Or 8xAA/16xAF .

Works for me, but I guess I am using the 61.40 drivers now.

 

[KimB]

Also, I was under the impression that it was VS 3.0 that was the big deal in SM3 so, as a non-techie person, I don't understand why there isn't anything utilizing it. Why did they put the effort into PS 3.0?

VS 3.0 is big on allowing new types of algorithms. PS 3.0 is mostly effiency improvements (though sometimes very significant). So the performance increases that FarCry is producing are what we should expect over the next 6-12 months from PS 3.0 implementations in games, though the differences are probably likely to become greater as time goes on.

 

[2senile]

Thanks Chalnoth.

So, basically, anyone interested in SM3 should forego this generation & get the nV5.x/ATi 500?

Also, with the 915G & RS400/RS480 chipsets comming out this year do you think SM3 is going to be held back?

 

[hmmm]

Other websites also now have the performance comparisons, and the results are similar.

http://www.techreport.com/etc/2004q3/farcry/index.x?pg=1

http://www.techreport.com/etc/2004q3/farcry/index.x?pg=1

Those disagree.

 

[KimB]

Doesn't look that way to me. Looking at the graphs, PS 3.0 still improves the performance, on average, of the 6800's, and still puts the XT and the Ultra at close performance, with the 6800 Ultra winning over the XT more often than not.

 

[2senile]

Also, with the 915G & RS400/RS480 chipsets comming out this year do you think SM3 is going to be held back?

Damn, I'm a sad bugger quoting myself!

The point I'm trying to make is that the above solutions are at the level of 5200/9200 performance & with a supposed 30% of all motherboards being equiped with onboard graphics I fear that games will continue to be coded for this "lowest common denominator".

 

[hmmm]

I have to say that, as long as there is no detrimental effect on other cards...

But there seems to be, hence...

X800 cards performance drops a little with the 1.2 patch

Which is why jvd said this:

I was allways under the impression that if a patch is a performance patch then it should increase performance for everyone . I wont be downloading this patch. no need for it and it will only slow me down

 

[ChrisW]

Not to mention the GFFX cards also take a hit with this patch.

 

[hmmm]

Doesn't look that way to me. Looking at the graphs, PS 3.0 still improves the performance, on average, of the 6800's, and still puts the XT and the Ultra at close performance, with the 6800 Ultra winning over the XT more often than not.

Um, no. You totally missed the point. Anand has crazy numbers because his tests didn't actually use AA. Someone said that was a dumb error for a supposedly professional site, especially when they make image quality claims. You said other places had similar numbers (that, like Anand's, showed *huge* gains; and the 6800 winning virtually *all* the time). I pointed out that you were wrong. The whole discussion was about Anand's anomolous numbers. That's the context for your comment. You are shifting ground.

I might also add that your new ground isn't that great either. TR's numbers show the XT PE winning the vast majority of AA+Ansio benchmarks at high resolutions (i.e. >1024x768).

 

[gordon]

Does this patch fix the graphical glitches that were present in the NV3x line of cards? I would assume it has given that the x-bit labs preview shows huge performance drops going from 1.1 to 1.2 for the 5950 Ultra (what else could be the reason for that?), but I haven't seen it explicitly noted anywhere or confirmed.

 

[Bad_Boy]

Does this patch fix the graphical glitches that were present in the NV3x line of cards?

yeah, nvnews has some pics of the banding being totally gone.

 

[Ante P]

Ante P, you really should do a geometric mean for tests like these. The test with 650 fps gets 50+ times the weight of a 12fps test, making the increase from 12 to 50 almost meaningless.

Also, when you do an overall average, you should calculate that from the overall result (i.e. 303/290 = 4.5%, except with a geometric mean), not an average of the individual percentages.

hehe I did the avarages a bit too quick
for the record this is a one hour article and that includes the actual benchmarking

as I said: quick and dirty

I did change the calculations now (avarage fps:es is still totally meaningless, but avarage percentage should make sense?)

oh well I have to go to work in half an hour so I can't fix anything yet anyways

btw if anyone has any more demos or anything (even the smallest app will do) where I can compare SM2.0 vs SM3.0 just let me know and I'll add it