Another game will soon support SM3.0

[Farid]

Link to homelan fed

Painkiller's futur add-on Battle Out Of Hell will support Shader Model 3.0.

Adrian Chmielarz, People can Fly's founder:

Adrian Chmielarz - By the time the add-on is released, many sweet games will be already out there, games that bring a lot of technology improvements (like DOOM 3 or HL2), so of course we’ll add our share of new stuff as well. We have hired a new programmer whose only task is to add new cool FX to the game. Definitely expect Pixel Shader 3.0 support.

 

[thop]

We have hired a new programmer whose only task is to add new cool FX to the game. Definitely expect Pixel Shader 3.0 support.

Hmmm interesting. I take it those sentences are totally unrelated but seem to be?

 

[jb]

It would be a safe bet that all TWIMTBP games will have SM3.0 sooner than a non-TWIMTBP game

But I thought the current version of PainKiller does not support PS2.0?

 

[RejZoR]

It does my friend it does Maybe its not obvious and it doesn't actually use it for everything,but its there,mixed with DX8 functions.

 

[London Geezer]

What impressed me the most of Painkiller (apart from the physics and the fact that it runs like a charm) is that no matter how close u get to a surface, the textures are always razor sharp, the detail texturing in this game is amazing, waaaaaaaayyyyyy better than anything i've seen in any other game.

 

[anaqer]

What impressed me the most of Painkiller (apart from the physics and the fact that it runs like a charm) is that no matter how close u get to a surface, the textures are always razor sharp, the detail texturing in this game is amazing, waaaaaaaayyyyyy better than anything i've seen in any other game.

True dat.
Last time I saw detail textures put to that good use was with Serious Sam - think it goes with the genre?

 

[Bouncing Zabaglione Bros.]

We don't know if it's SM3.0 on the headline with SM2.0 in the small print like Far Cry. It seems the developers make a lot of noise about SM3.0 in order to get Nvidia's cash, and then do SM2.0 anyway for the vast majority of their customers.

 

[Hellbinder]

In other Words expect some PS2.0B compatible Additions that are Only available through Id detected Nvidia cards... But have a SM3 "label" on them.. Oh and they also use FP16 even though that is totally illegal for SM3.0..

Its ok though becuase Nvidia will continue to use their PR machine and wads of cash to cover up the truth and mislead people.

 

[Razor1]

Depends on the shader Hellbinder, you can use 2.0 and 3.0 simultaneously. No problem with that at all. If the shader doesn't require fp 32 no need to do it. Its not a whole lot slower anyways, but thats not that point use it if its only needed.

 

[Ostsol]

In other Words expect some PS2.0B compatible Additions that are Only available through Id detected Nvidia cards... But have a SM3 "label" on them.. Oh and they also use FP16 even though that is totally illegal for SM3.0..

Its ok though becuase Nvidia will continue to use their PR machine and wads of cash to cover up the truth and mislead people.

AFAIK, precision hints still okay in PS3.0. The HLSL spec has a half-float datatype, too. The only thing that would be illegal is if the drivers or video card didn't support FP32. In any case, I really doubt that a developer would bother to use PS3.0 unless PS2.0 simply didn't suffice (and I'm not talking about precision or instruction counts, of course).

 

[Demirug]

The new SDK contains information about SM3 and precision.

Shader Model 3

Specifying Full or Partial Precision

ps_3_0 (and ps_2_x) provides support for 2 levels of precision:


ps_3_0
Full fp32 or higher
Partial precision fp16=s10e5

ps_2_0
Full fp24=s16e7 or higher
Partial precision fp16=s10e5

ps_3_0 supports more presision than ps_2_0 does. By default all operations occur at the full precision level.

Partial precision (see Modifiers for ps_2_0 and Above) is requested by adding the _pp modifier to shader code (provided that the underlying implementation supports it). Implementations are always free to ignore the modifier and perform the affected operations in full precision.

 

[Demirug]

From a techninal point of view it isn't a bad thing to use the compiler profile that fit the target chip best. This is one of the advantages if you use a high level language.

I am agree that it is a unpleasant behavior to abuse this technical reality for marketing purposes. Maybe it is only me but during the time ATI plays the SM2 card (like HL2 benchs) this was not a big topic anyway.

 

[Xmas]

Perhaps it might be better to wait until the add-on is out before drawing any premature conclusions.

 

[overclocked]

As someone said i didn´t know about PK and when i played it, it blow me away, soo good, real Action.. 8)

 

[micron]

OK..... I've seen Hellbinder go on this 'rant' numerous times now, and I would like to get to the bottom of this.
Is fp16 'not part of' sm3.0?

 

[tEd]

OK..... I've seen Hellbinder go on this 'rant' numerous times now, and I would like to get to the bottom of this.
Is fp16 'not part of' sm3.0?

see demirugs post above

 

[jvd]

OK..... I've seen Hellbinder go on this 'rant' numerous times now, and I would like to get to the bottom of this.
Is fp16 'not part of' sm3.0?

It is . But only if a program says it can be used.

If the code is over rided to use fp16 instead of fp32 when its called for then it is not part of dx 9 . But nvidia wouldn't do that would they ?

 

[Reverend]

Oh and they also use FP16 even though that is totally illegal for SM3.0

Show me. The "illegal" part, I mean.

 

[Dave Baumann]

OK..... I've seen Hellbinder go on this 'rant' numerous times now, and I would like to get to the bottom of this.
Is fp16 'not part of' sm3.0?

No, thats bullshit.

 

[digitalwanderer]

I dunno, why do they have TeamDX police if it isn't illegal then? :|