[HardOCP] Nvidia cards much faster than AMD cards in "Rage" game.

[tviceman]

Hard ocp sucks.

 

[WMD]

16X AF creates broken lines and any AA above 4X blurs out the textures. Set to 4XAF 4XAA it looks great! I am using ATI and no performance issues here.

 

[toyota]

I don't mean to rag on HardOCP unnecessarily, but they're idiots who didn't do their research.

http://www.hardocp.com/image.html?image=MTMxODgyNTI4NmQwb1lRSUFnU1JfNl8xM19sLmpwZw==

That corruption happens because they forced 16x AF in their drivers. Tech5 can't handle forced AF, it doesn't work across texture tiles. It also has some pretty serious performance repercussions at times as a result.

yep I mentioned that in the thread over there. sometimes its amazing that they miss such obvious stuff.

 

[happy medium]

CUDA GPU transcoding is GPU assisted transcoding of the heavily compressed megatextures used in Rage basically.

Anandtech had a bit of a write up about it:



Source..

ID also tried to use OpenCL to do the same thing on the Radeons, but the performance apparently wasn't good enough to justify including it in the final release.

Thanks, for the reply.
I learned something .:thumbsup:

 

[blackened23]

H is not biased, their reviews are top notch. It is true that NV cards are much faster in Rage than AMDs. You won't see it if you test at low res or low AA, because Rage has a 60 fps max limit.

But so what, NV cards are a lot faster than AMD in LP2, Hawx 2, Crysis 2 etc..

There's a theme here. You may know what it is.

I've yet to find a system that doens't play Rage extremely well. Even slightly lower end cards (ie single 6950) play it ridiculously well at 1080p.

Anyway, I'm in shock that they enabled post processed effects in the driver for a benchmark, thats pretty amateur. You'd think they would know better? Most people are aware, you should never do this because the performance penalty for driver effects is severe compared to ingame IQ effects. You should always use "let application decide" because the driver applies effects to an entire viewable area, but an ingame option applies it where its needed - the performance penalty for using a driver IQ effect can be severe. A good example that comes to mind is Dead Space. If you enable 16Q AA in nvidia inspector, the game slows horribly in areas with lighting....this is despite the fact that Dead Space isn't graphically demanding at all.

 

[happy medium]

That Cuda Gpu texturing must really help..

It looks like AMD's minimums are really bad.

 

[3DVagabond]

Geez... I think that we need a third major video card manufacturer, just so the ATI fanboys and Nvidia fanboys both have something to rally against.

Oh, and I mean a Competitive video card manufacturer, not Intel or Matrox!

You know what they say, "haters are gonna hate". There's no easy way to stop it. Might as well sit back, eat your popcorn, and be entertained.

 

[Madcatatlas]

Hard ocp sucks.

Seconded.

they get paid to do galaxy reviews

 

[BFG10K]

Most people are aware, you should never do this because the performance penalty for driver effects is severe compared to ingame IQ effects. You should always use "let application decide" because the driver applies effects to an entire viewable area, but an ingame option applies it where its needed - the performance penalty for using a driver IQ effect can be severe.

This blanket statement is not always correct. I’ve seen numerous instances where driver controls produced better performance and/or IQ than in-game settings.

A good example that comes to mind is Dead Space. If you enable 16Q AA in nvidia inspector, the game slows horribly in areas with lighting....this is despite the fact that Dead Space isn't graphically demanding at all.

You might have a point if Dead Space has an in-game 16xQ setting to compare with. Does it?

 

[blackened23]

This blanket statement is not always correct. I’ve seen numerous instances where driver controls produced better performance and/or IQ than in-game settings.

Pretty sure it is correct. Do you understand how the driver applies IQ settings versus how a game does it? The driver IQ setting does MUCH more work to accomplish the same thing. The driver method is always slower, period. So I take it you're one of the ones who apply IQ settings through the driver?

If you use the ingame AA option, the developer decides which parts of the rendering process would actually benefit from it and which would not. If you do it with the driver it applies to the entire scene and in areas where it doesn't make sense. If you do this in a modern game (or not so modern game, as in dead space' case) you will murder your framerate, period. I would entertain you with benchmarks if I cared more, as I have tested this numerous times in the past -- the driver IQ setting is without exception always slower than an ingame option for the same thing. Its not hard to understand really. Optimized rendering where its needed, versus something applied to an entire scene. See, not hard to understand Maybe i'll post screenshots to prove this. Shrug.

You might have a point if Dead Space has an in-game 16xQ setting to compare with. Does it?

Nah bro you tell me, do you think Dead Space 1 is graphically demanding enough to dip to 30 fps? Because it does dip that low in heavily lighted areas despite looking like a game using the half life engine. On a GTX 580.

Actually, while i'm at it I just tested 8x AA in the nvidia inspector and the game dipped to 50 fps in the heavily lighted scene where they approach the IG ishamura for the first time. It also takes a nosedive to 25 fps in areas where you go to zero gravity. (when you're transitioning and walking through a zero gravity door for the first time) Now you tell me, do you think a game that looks like half life 2 at release _really_ wouldn't do this better with an ingame option?

 

[BFG10K]

Pretty sure it is correct.

It isn’t. Two examples off the top of my head:

Fear 1: in-game AF doesn’t control trilinear properly leading to visible mip-banding, driver AF fixes it.

Jericho: in-game AA runs slower AND looks inferior to driver AA.

There are many more, such as games using deferred rendering and/or HDR paths where their AA settings have no effect in these modes, but driver AA keeps working. Or games that are restricted to lower levels of AF/AA while the hardware is capable of going higher (e.g. super-sampling).

Do you understand how the driver applies IQ settings versus how a game does it? The driver IQ setting does MUCH more work to accomplish the same thing. The driver method is always slower, period.

You’re making blanket statements which aren’t accurate. If the driver always behaves the same way, why do we have AA flags? Do you understand what application specific optimizations are?

So I take it you're one of the ones who apply IQ settings through the driver?

It depends on the game and the setting. Certainly, one would be foolish to stick to only one option given different games behave differently.

Nah bro you tell me, do you think Dead Space 1 is graphically demanding enough to dip to 30 fps? Because it does dip that low in heavily lighted areas despite looking like a game using the half life engine. On a GTX 580.

What’s your point of reference to compare? No AA? That’s a logical fallacy.

The fact is you have no valid point of reference because you can’t set 16xQ in-game. If you could and analyzed IQ/performance which demonstrated in-game works better, then you’d have a point.

Actually, while i'm at it I just tested 8x AA in the nvidia inspector and the game dipped the 30 fps in the heavily lighted scene where they approach the IG ishamura. It also takes a nosedive to 25 fps in areas where you go to zero gravity. Now you tell me, do you think a game that looks like half life 2 at release _really_ wouldn't do this better with an ingame option?

That’s great but it doesn’t prove anything other than it’s costly to run MSAA on a deferred render while filtering all polygon edges. But we already knew that.

The question is, would an in-game setting produce better IQ and run faster? Like I demonstrated above, not necessarily, given I can cite numerous examples where it doesn’t.

 

[SirPauly]

I agree with BFG10K.

 

[Silverforce11]

So they provided an example of a glitch in the Rage renderer.. great, now people know not to force quality IQ settings through the driver.

Really, it was a good bench review, clearly shows NV being superior, especially with their CUDA texture transcoding.

Ultimately, Rage is a crap game that will be forgotten very soon, similar to Crysis 2. Both have terrible sales on the PC, and their devs will use the poor sales # to once again claim pc gaming is dead blah blah. You know its bad when even Carmack thinks its pointless to develop games for the PC.

 

[tviceman]

So they provided an example of a glitch in the Rage renderer.. great, now people know not to force quality IQ settings through the driver.

Really, it was a good bench review, clearly shows NV being superior, especially with their CUDA texture transcoding.

Ultimately, Rage is a crap game that will be forgotten very soon, similar to Crysis 2. Both have terrible sales on the PC, and their devs will use the poor sales # to once again claim pc gaming is dead blah blah. You know its bad when even Carmack thinks its pointless to develop games for the PC.

I'm sure AMD will have near perfect performance with their next driver release. No one knows how Rage is selling on PC, you are extremely exaggerating Carmack's comments on Rage development, and I personally do not think Rage is crap.

 

[SirPauly]

So they provided an example of a glitch in the Rage renderer.. great, now people know not to force quality IQ settings through the driver.

Really, it was a good bench review, clearly shows NV being superior, especially with their CUDA texture transcoding.

Ultimately, Rage is a crap game that will be forgotten very soon, similar to Crysis 2. Both have terrible sales on the PC, and their devs will use the poor sales # to once again claim pc gaming is dead blah blah. You know its bad when even Carmack thinks its pointless to develop games for the PC.

I see nothing but growth potential with the PC platform. Crysis 2 isn't a crap game and is actually fun and had nice reviews and sold over 3 million copies with around 330, 000 retail boxed editions sold for the PC and probably around 200,000 digital copies sold for the PC, so maybe around 500,000+ copies sold.

Rage has sold over 70,000 retail boxed editions and probably around 45,000 digital copies.

Deus Ex, another nice game has sold around 170,000 copies for the PC with retailed boxed editions and probably around 100,000 sold for digital copies.

Is it because nVidia may have a hardware/software advantage -- instantly a crap game?

 

[(sic)Klown12]

No one knows how Rage is selling on PC

We do know some numbers from the first week, and it's not looking all that well.
http://www.msxbox-world.com/news/ar...f-million-in-first-week-new-ips-not-dead.html

There's a fear that bringing something new to gaming is filled with much risk as gamers are quite the fickle bunch. But as with the recent Dead Island and now RAGE it seems new IPs are making waves amongst gaming communities.


RAGE released recently under the wings of publisher Bethesda, and it's now suggested by chart tracking data from VGChartz that RAGE sold well over half a million units in its first week, which must be encouraging for developer id Software.

The estimated data suggests that in week 1 RAGE sold just over 400K units on Xbox 360 and a further 240K units on PS3 (with some 60K on traceable PC platforms). That's great for a new IP and hopefully means that over time the game sells well over the 1 million mark if it hasn't already. The figures certainly prove that providing the content gamers are after and a bit of clever marketing can go a long way in making new ideas reach out to gamers and their spending habits.

 

[notty22]

I plan on picking it up eventually. If you think every game is crap, blah blah, it's time to give up gaming and get a different hobby. It's a fact of business life, that games have to be co-developed for pc/consoles. I just got Forza 4 for xbox 360, and it rocks.
I wish I could get a PC version of that !

 

[blackened23]

We do know some numbers from the first week, and it's not looking all that well.
http://www.msxbox-world.com/news/ar...f-million-in-first-week-new-ips-not-dead.html

Sarcasm? I'd say near a million in sales the first week isn't a disappointment. 300k ps3, nearly 500k xbox 360, yeah, that sounds pretty damn good to me.

Not that I like the game. I'd say wait for a steam sale, because for a 10 hour single player experience I wasn't impressed. Thats just me though. You can easily get better
single player entertainment through batman: AA, dead space 1/2, among others.

 

[SirPauly]

Sarcasm? I'd say near a million in sales the first week isn't a disappointment. 300k ps3, nearly 500k xbox 360, yeah, that sounds pretty damn good to me.

I agree, looks like a damn good seller.

 

[happy medium]

Is it because nVidia may have a hardware/software advantage -- instantly a crap game?

The reviews seem to paint a different picture,it seems to be a good game, not great but good.

 

[TakeNoPrisoners]

Those screens look really good!

Judging by those screens it beats the original Crysis!

 

[Silverforce11]

When i said the sales # are bad, i am talking about the PC sales. RAGE getting over 500k, but almost all from consoles. The PC # is 60k, which is pretty damn bad. Crysis 2 being released long ago only managing less than half a million PC sales (a lot from discounted digital downloads) is very low for a AAA budget title. Crysis 1 sold much better.

BC2: Developed on the PC, ported to consoles. In its first year, sold nearly 9M copies. ~2.5M PS3, ~3M 360 and >3M PC. PC sales are still continuing via digital dl. It did a lot better than BC1.

 

[(sic)Klown12]

Sarcasm? I'd say near a million in sales the first week isn't a disappointment. 300k ps3, nearly 500k xbox 360, yeah, that sounds pretty damn good to me.

Not that I like the game. I'd say wait for a steam sale, because for a 10 hour single player experience I wasn't impressed. Thats just me though. You can easily get better
single player entertainment through batman: AA, dead space 1/2, among others.

I'm talking specifically the numbers for the PC. With all the hype building up prior to release, 60K is pretty small. With all the bad press that has followed, I doubt sales numbers will pick up any steam outside of some huge price cut.

 

[Borealis7]

Wait...so what happens to Rage on a 120Hz monitor? the renderer is capped at 60FPS!

 

[notty22]

Anyone have a link for this 60,000 PC number. I have dozens of links to business articles how there are very little to no reliable pc sales #'s released. Unless a developer eventually releases numbers that are used for effect their IE- quarterly statements.
By the way the first Crysis also had poor first month sales, which was announced after the fact, because sales picked up so much after that point.