Why was Far Cry 1.2 really recalled?

[tornadotsunamilife]

I doubt that we will see that hardware for a while. If NV40 can't run something like 3dmark03 Game Test 4 in full FP32 at the same speed as FP24, IMO it won't be until at least NV60/R600 that we will see cards that can run FP32 the same speed as a FP16/FP32 mix in a game as complex as Unreal 3. Plus with the way NV40 is designed, Nvidia really isn't losing any FP32 performance by offering FP16, as the FP16 shader hardware is combined to get FP32. If NV50 is designed the same way and R520 isn't, I think we will see Nvidia take the lead in intensive DX9 apps like Unreal 3.

i doubt we will, since i personally believe there will be some form of sm4 in effect. however i may be wrong and since not much has been mentioned of dx10 i will just have to wait and see. (im guessing R6X0/NV6x cores will be dx10 capable, that is to say use dx10 eye candy)

 

[tranCendenZ]

a. changing the VendorID/DeviceID from X800 to GF 6800
b. Wrapping the 3.0 shaders to 2.0b shaders.

You gain as much as a 20% performance boost in FarCry 1.2!!!

Now what the heck is going on here? Obviously something underhanded in regards to Nvidia optimizations that has nothing to do with your actual hardware!?

Nothing underhanded, it's called one pass lighting. The 1.2 patch doesn't even work properly with ATI cards in the 2.0 path, you expect the 2.0b path to work great? Plus the 3.0 path is still in beta and can't even be enabed yet. By the time its finished, odds are there will be plenty of stuff that can't be done in sm2.0. You can't expect developers to develop for old technology forever.

 

[tranCendenZ]

if everything in sm3.0 can be done or converted to sm 2.0b what was the point of sm 3.0? Nvidia would have had to know this and if that was the case i dont see why they would push the 3.0 so much in all their marketing.

Everything can't be done in 2.0b that is available in 3.0, that is BS. On page 3 of this thread you can see all the deficiencies of 2.0b compared to 3.0.

It can be visually emulated in many more passes, though.

The point of SM3.0 is to allow for more complex and efficient shaders. This is why Microsoft developed the standard. It has not yet been used to its fullest in the FarCry 1.2 patch.

ATI is running a SM3.0/partial precision FUD campaign currently because of what happened with Doom3. Honestly I probably would too if I lost that badly in the benchmarks.

 

[tranCendenZ]

i doubt we will, since i personally believe there will be some form of sm4 in effect. however i may be wrong and since not much has been mentioned of dx10 i will just have to wait and see. (im guessing R6X0/NV6x cores will be dx10 capable, that is to say use dx10 eye candy)

If SM3 won't be out until this august and the NV40 is the only card that supports it, I seriously doubt we will see SM4 until Longhorn (i.e. not next year).

 

[lopoetve]

This is my bone of contention with Nvidia.... Nvidia is marketing to everyone in the freaking world how they are compliant yet instruct TWIMTBP partners to mix FP16/32. What this is telling me is they rushed half assed support for SM3.0 knowing they couldn't fully support the minimum requirements without turning into a slide show.
.

You have no proof of this. Like I said before, optomizing for increased performance where it doesn't effect visual quality is always a good thing.

 

[lopoetve]

Nice way to talk to someone on this forum. The mod ban for less these days.

You obviously haven't been here long. Insulting what someone says is different than insulting them, and I wasn't even doing that!

 

[lopoetve]

I trust them much more than Toms!

One form of shit vs. another. Pointless.

As for the rest, it's still unfounded bs.

 

[lopoetve]

It makes sense to people trying to justify their purchase, but I'd feel lied too.

The fact still remains it can not support running FP32 by it self. Why bother pushing something so big only to down grade the output because it can't render it at a reasonable frame rate? Maybe the marketing should go as such: "We support "partial" SM3.0".

Like I said on the first post, something big is going to happen very soon.

Why would you feel lied to? The Developers choose to use lower precision, no one else. The original GF256 wasn't super great at T&L, but it was still a usable feature.

And you're still talking about "downgrading the output".

THERE IS NO REDUCTION IN VISUAL QUALITY, the developers said so (developers >>>>>>>>>>>>>>>> driverheaven), AND it's just to increase speed for more headroom/buffer (developers >>>>>driverheaven bs editorial). There is NO proof (Because it doesn't exist) that the cards can't run at full 32 the whole time.

 

[CIWS]

Nice way to talk to someone on this forum. The mod ban for less these days.

This is not true, please do not present it as such.

Just because he asks wtf are you talking about does not violate the rules of this forum. Being insulting to them, Flaming them, Crapping on their threads, being a Troll, etc does and it's this people are being Banned for, and to be quite honest I've personally given several folks here more chances than was required. However the warning posted at the top of this forum about namecalling is very clear. If you choose to ignore it, that's all it can take. Most people usually only see the last instance of something that got a person Banned. What you usually don't see is the multiple instances before that that led up to that point.

 

[@trapine]

ATI is running a SM3.0/partial precision FUD campaign currently because of what happened with Doom3. Honestly I probably would too if I lost that badly in the benchmarks

Sorry but its Nvidia thats running a pretty damed good FUD campain With Them tryin to spread FUD of 20% or more performance in games with Smoothmarketing3 ATI look to be going the right way about it to me.I.E impliment the tech when the Card has the power to use it?
Everyone has been bouncing around sprouting SM3 is coming on heaps of up and coming titles and its gonna give good performance boosts?But to date its just a flop and hotair IT will be good BUT when the Tech has the power to run it properly?I>E NV50 and R500 probably?Not NV40!

 

[tranCendenZ]

Sorry but its Nvidia thats running a pretty damed good FUD campain With Them tryin to spread FUD of 20% or more performance in games with Smoothmarketing3 ATI look to be going the right way about it to me.I.E impliment the tech when the Card has the power to use it?

More FUD. Nvidia cards are more than powerful enough to run FP32, and there is proof to back that up. FarCry FP16 vs FP32 testing, 3dmark03 results, homebrew apps, etc. Still, FP16 can give a performance boost with no IQ loss. Stop making things up.

As benchmarked by Chrisray of Beyond3d/Nvnews, the difference between full FP32 mode and full FP16 on 6800 cards is 5-10% in FarCry depending on the scene, and there is little IQ difference even when using full FP16 and no FP32. The reason FP16 is used in some shaders is because there would be *no benefit to using FP32* on some shaders.

Everyone has been bouncing around sprouting SM3 is coming on heaps of up and coming titles and its gonna give good performance boosts?But to date its just a flop and hotair IT will be good BUT when the Tech has the power to run it properly?I>E NV50 and R500 probably?Not NV40!

NV40 is running it "properly" right now - if you look at actual FP16 -> FP32 performance numbers and IQ differences along with the SM3.0 specs instead of Driverheaven BS you could see this. We are just seeing the beginning of SM3.0 support.

 

[burningrave101]

Sorry but its Nvidia thats running a pretty damed good FUD campain With Them tryin to spread FUD of 20% or more performance in games with Smoothmarketing3 ATI look to be going the right way about it to me.I.E impliment the tech when the Card has the power to use it?
Everyone has been bouncing around sprouting SM3 is coming on heaps of up and coming titles and its gonna give good performance boosts?But to date its just a flop and hotair IT will be good BUT when the Tech has the power to run it properly?I>E NV50 and R500 probably?Not NV40!

And to date it hasn't? rofl

You and the rest of the ATI community seem to think you know alot about SM3. ATI users try to show that there is almost no difference between PS 2.0b and PS 3.0 and your all so wrong its not even funny. And how would YOU know there isn't 20% performance gains to come from SM 3.0? There were 20% performance gains in certain parts of Far Cry already lol. You've seen SM 3.0 implemented into ONE game by Crytek. Thats it. So now that you you've seen it in ONE game you know how its going to perform for the next TWO years in dozens of other games? Give me a break.

You have NO proof that the NV40 cores dont have enough power to run it.

And Crytek doesn't have the worlds greatest programmers in the world if you get my drift.

ATI knows this and that is why they are spreading FUD about SM 3.0 and FP16/FP32 precision.

While NVIDIA is loudly and proudly advertising the new shader model, ATi is attempting to downplay it. A wonderful and quite humorous example of this tactic is a developer presentation titled "Save the Nanosecond" by ATIs Richard Huddy which was accidentally leaked on the internet. The trouble was, the presentation still sported some personal notes which made it clear that the presenter was trying to convince developers to stay away from Flow Control in PS3.0, as it incurs a very tangible performance hit. Quote:

"Steer people away from flow control in ps3.0 because we expect it to hurt badly. [Also it's the main extra feature on NV40 vs R420 so let's discourage people from using it until R5xx shows up with decent performance...]"

Internet forums everywhere are still tied up in all sorts of debates about how this statement should best be interpreted. One way of reading this quote would be that ATi's upcoming chips are either still too slow for this feature or simply aren't designed for it. Another would be that it won't be until the generation of cards following the one being launched now, that graphics accelerators will be able to cope with the complexity caused by this feature...

http://graphics.tomshardware.com/graphic/20040723/his-08.html

 

[tranCendenZ]

"Steer people away from flow control in ps3.0 because we expect it to hurt badly. [Also it's the main extra feature on NV40 vs R420 so let's discourage people from using it until R5xx shows up with decent performance...]"

lol that is classic

ATI's new presentation "Steer people away from OpenGL"

 

[^eMpTy^]

Sorry but its Nvidia thats running a pretty damed good FUD campain With Them tryin to spread FUD of 20% or more performance in games with Smoothmarketing3 ATI look to be going the right way about it to me.I.E impliment the tech when the Card has the power to use it?
Everyone has been bouncing around sprouting SM3 is coming on heaps of up and coming titles and its gonna give good performance boosts?But to date its just a flop and hotair IT will be good BUT when the Tech has the power to run it properly?I>E NV50 and R500 probably?Not NV40!

I get the feeling everyone is going over to teamradeon, reading all thier bs...then rolling back over here like they know something we don't...

Congratulations suckers, you all just god duped by some really silly press from a company that doesn't have the self-respect to admit defeat in doom3 and carry on with dignity...

You can say it as many times as you want but it won't change the truth...even if everything ATi is saying right now is true (which it's seriously all just a sad sad attempt at damage control with the doom3 benchmarks), they still only have one card available on the market and it's the castrated X800 Pro which sucks even when comparing its native FP24 performance to Nvidia's FP32...

But sure, there's some wierd FarCry numbers running around...and I'm sure that's confusing...but let's do this...let's wait till the patch comes back out again and get the word straight from the horses mouth about 2.0b and 3.0 performance instead of jumping to conclusions...

 

[^eMpTy^]

I hope im not going off track here but i want to ask a question. All the benchies we have seen are with nvidias bi/tri optimized filtering off where on the oposite end ATI's optimizations are on corroect? had the Nvidia hardware been benched with its optimizations on it would gain a performance boost correct? (thats the whole point of an optimization). I agree with someone who posted before, as long as i dont notice the difference i really dont care.

Some benchmarks had Nvidia's optimizations on, others didn't...most turned them on to be fair since you can't turn ATi's off.

if everything in sm3.0 can be done or converted to sm 2.0b what was the point of sm 3.0? Nvidia would have had to know this and if that was the case i dont see why they would push the 3.0 so much in all their marketing.

I'm really not sure about this right now. Clearly in FarCry there is something going on in the 2.0b path but I don't know why they haven't officially announced it yet. Either way Nvidia is hyping it for the same reason anyone hypes anything, they have it, ATi doesn't.

i think there is still a lot more headroom for nivida hardware as far as performance is concerned because it is a new artitecture and needs to be polished more and more where as ATI's drivers appear to be already touching their full potential. (there wasnt much of a performance increase from 4.6-4.7 was there?)

There was a modest perforance improvement, but ATi also removed the 4.7 catalysts from the 3dmark03 program due to artifacts that started showing up. It seems to me they are trying to squeeze out every drop of performance and they're really starting to cut corners.

anyway to close i hope that we get some of these things cleared up before i get my new videocard in hand.

ati's performance seems to be a tad better in D3D with the new x800xt. The main reason i have a BFG 6800 ultra OC on order is because with SM 3.0 it appears to make up that difference. If ATI hardware can pull the same performance increase with 2.0b then ill be really upset about this 500 dollar purchase because from all the marketing i was thinking that ATI couldnt take advantage of this in any way. If it can, then i probably would sway my purchasing decision back on the other side of the fence. I just hope some official word comes out on this in the very near future(something other than the driver heaven article). that way i can make the right decision and be safe to not need to upgrade for the next 8-12 months.

Emphasis on the word "tad", the Ultra and XTPE are effectively equals in D3D, everyone pretty much agrees on this point. PS30 is just starting to get off the ground, chances are you'll see a lot more improvements from it later on down the road. Keep in mind that in the past Driver Heaven has been wrong about several things they've accused Nvidia of...from what I've seen they are extremely ATi biased, not to mention they don't really know what the hell they are talking about. As pointed out aleady, they don't even understand what the names FP16 FP24 and FP32 mean.

 

[@trapine]

But sure, there's some wierd FarCry numbers running around...and I'm sure that's confusing...but let's do this...let's wait till the patch comes back out again and get the word straight from the horses mouth about 2.0b and 3.0 performance instead of jumping to conclusions

You do realise you just contradicted the hell out of your self then On the one hand you are saying the above statement>But you are also then going to accept as gospel the Doom3 Numbers.When #1 Doom3 is not out yet.X800XT only got beat by 6800GT by 2.8FPS and by 7.6 with 6800U.Very easy numbers to make up?And who knows how its gonna bench?
So you are getting just as carried away as the people you are trying to shoot down ^eMpTy^
Take a bit of your own advice and just wait till Doom3 gets here?

 

[^eMpTy^]

You do realise you just contradicted the hell out of your self then On the one hand you are saying the above statement>But you are also then going to accept as gospel the Doom3 Numbers.When #1 Doom3 is not out yet.X800XT only got beat by 6800GT by 2.8FPS and by 7.6 with 6800U.Very easy numbers to make up?And who knows how its gonna bench?
So you are getting just a carried away as the people you are trying to shoot down ^eMpTy^
Take a bit of your own advice and just wait till Doom3 gets here?

There's a big ass difference between official numbers straight from the developer of Doom3 after it has already gone gold and numbers from random ATi people messing around with their product ID and enabling unsupported features in a game to get an as yet unproven performance boost. Not to mention the patch just got recalled because it was crashing on ATi hardware.

Any time the GT beats the XTPE in any tier one game, that's a big loss in my book. You can call it "only 2.8fps" all day long...in Doom3 every fps is going to count and if you wanna dismiss those numbers as irrelevent you'll be lonely cuz everyone else thinks they are pretty damned important.

If you ever find yourself thinking that I contradicted myself, think twice, because it happens very rarely.

 

[Xenozx]

kool thanks for the info ^empty^

 

[ZenOps]

I wouldn't say there is no reduction in quality. Just like ATi's optimized trilinear there is a reduction in quality. Its just a matter of how noticable it will be. If the visual differences are slight then its valid IMO. Although I can imagine it would be a SOB to code every object as 32 or 16FP or even automatically distance based code, without having a visual anomaly every so often.

ATi is at a disadvantage, in any game that uses partial precision. ATi does 24bitFP on all objects no matter how close or how far, or how simple or complex and object is. If ATi allowed partial precision it would probably be 12bitFP, for things like the sky textures on a game that is entirely located on the ground as you do not need high-precision on things that are far from the users perspective (Nvidia was even talking about 12bitINT in some situations)

Ps: I'd almost say you could use 6bitFP for the sky in a lot of games. It really doesn't matter if the clouds are programmed to appear to the user to be 11kilometers up or 12kilometers up, it will look approximately the same regardless of any semblance of precision. However, screw up the precision on a crate lying on the ground by 1/10th of a millemeter, and I'd bet someone will find the texture/shimmer seam every time.

 

[tranCendenZ]

I wouldn't say there is no reduction in quality. Just like ATi's optimized trilinear there is a reduction in quality. Its just a matter of how noticable it will be. If the visual differences are slight then its valid IMO. Although I can imagine it would be a SOB to code every object as 32 or 16FP or even automatically distance based code, without having a visual anomaly every so often.

No, that is incorrect. There is no IQ loss at all with FP16 if a developer properly assesses the shader they will be attempting to use FP16 on. If precision errors are found when using FP16 with a shader, the developer can then use highest FP32 precision on Nvidia cards instead.

 

[^eMpTy^]

No, that is incorrect. There is no IQ loss at all with FP16 if a developer properly assesses the shader they will be attempting to use FP16 on. If precision errors are found when using FP16 with a shader, the developer can then use highest FP32 precision on Nvidia cards instead.

Exactly. This isn't a matter of trading iq for performance. If the shader can operate at 16bit, there is no iq loss from 32 bit...the concept here is that fp32 literally is not needed at all for some shaders...

 

[lopoetve]

I wouldn't say there is no reduction in quality. Just like ATi's optimized trilinear there is a reduction in quality. Its just a matter of how noticable it will be. If the visual differences are slight then its valid IMO. Although I can imagine it would be a SOB to code every object as 32 or 16FP or even automatically distance based code, without having a visual anomaly every so often.

ATi is at a disadvantage, in any game that uses partial precision. ATi does 24bitFP on all objects no matter how close or how far, or how simple or complex and object is. If ATi allowed partial precision it would probably be 12bitFP, for things like the sky textures on a game that is entirely located on the ground as you do not need high-precision on things that are far from the users perspective (Nvidia was even talking about 12bitINT in some situations)

Ps: I'd almost say you could use 6bitFP for the sky in a lot of games. It really doesn't matter if the clouds are programmed to appear to the user to be 11kilometers up or 12kilometers up, it will look approximately the same regardless of any semblance of precision. However, screw up the precision on a crate lying on the ground by 1/10th of a millemeter, and I'd bet someone will find the texture/shimmer seam every time.

The FP16 shaders are supposed to be on things that you can't see, thus the performance increase, and no (according to the developers) decrease in IQ.

 

[lopoetve]

Exactly. This isn't a matter of trading iq for performance. If the shader can operate at 16bit, there is no iq loss from 32 bit...the concept here is that fp32 literally is not needed at all for some shaders...

Bingo. Sometimes, you don't need all those zeros, or the overlap of another shader renders this one moot, BUT you still have to render it.

 

[tornadotsunamilife]

If SM3 won't be out until this august and the NV40 is the only card that supports it, I seriously doubt we will see SM4 until Longhorn (i.e. not next year).

that's what i mean! dx10 will come with longhorn

 

[KakimotoR]

Why are we even debating and fighting for NVIDIA and ATi? They dun even pay us money!
There is no point comparing in my opinion.

 

[avatar_of_might]

Why are we even debating and fighting for NVIDIA and ATi? They dun even pay us money!
There is no point comparing in my opinion.

very well said!!! i agree completely.
im very happy with the 1.1 patch so no big hurry on the 1.2 patch

 

[Stereophile]

What I find really annoying is the official servers are still running the recalled patch !!!
wtf

So I can play on the official servers or unofficial but not both.
Which really sucks cause from my location it's hard to even find a semi populated server with a ping less than 100.