RenderMonkey-Is this ATI's "Cg"?
- Thread starter ben6
- Start date
Doomtrooper
Veteran
Doomtrooper
Veteran
Geeforcer said:
I think some company forced THEIR HAND didn't it..ATI is going to sit back and watch Nvidia seduce all the developers with their own compiler..NO. In fact if you go back through that thread you will see I posted that this would happen, NvidiaCG, ATI CG, Matrox CG, then there is some people that think Nvidia released CG to be 'nice' to game developers.
Now its a battle of the compilers, my compiler is better than yours...just what we needed more confusion for developers.
Initially was impressed with whwat Cg represented then it became obvious that NVIDIA created it with the intent of keeping Cg close to itself.
Now ATi responds with an 'anything you can do, we can do better' stance.
All I will say is... OoOoOoPsSsSs
But what real choice did they have?
Now ATi responds with an 'anything you can do, we can do better' stance.
All I will say is... OoOoOoPsSsSs
But what real choice did they have?
Who said there was going to be a Matrox Cg (hrm maybe I shouldn't post that but ah well...) Irregardless, it's a little early to say this is a compiler ala Cg. It makes a lot of sense for ATI to release something like this , since NVIDIA's Cg wouldn't support PS1.4 anyway except through support of PS2.0 in the fall with DX9.
Doomtrooper
Veteran
I never stated there was going to be a Matrox CG , what I posted in the other mile long CG thread that ATI WOULD not make a profile for Nvidia CG, they would have to release their own compiler, then Matrox would be forced to do the same then SIS then 3Dlabs.
I think Misae states it best..
I think Misae states it best..
Doomtrooper said:I think some company forced THEIR HAND didn't it..ATI is going to sit back and watch Nvidia seduce all the developers with their own compiler..NO. In fact if you go back through that thread you will see I posted that this would happen, NvidiaCG, ATI CG, Matrox CG, then there is some people that think Nvidia released CG to be 'nice' to game developers.
Now its a battle of the compilers, my compiler is better than yours...just what we needed more confusion for developers.
Seduce developers with CG? But I thought that CG was a horrible thing that no one needed or was going to use
Anyway, we don't even know whether RM contains an HLSL compiler or not, but one this is certain - it is a advanced tools package and I remember quite clearly that you spoke against any IHV-developed tools.Doomtrooper
Veteran
Geeforcer said:Seduce developers with CG? But I thought that CG was a horrible thing that no one needed or was going to use
I spoke strongly against optimized code for certain hardware, which Nvidia CG did, exposed more optimizations ONLY for Nvidia hardware.
With a name like yours it's not a big deal, but for people that like choice then it is
We don't know if this meets the same licensing restriction criteria that Cg seems like it will. The wording doesn't look like it, and it is terse and without marketing buzzwords, as I'd expect for a siggraph presentation description. Can't even be sure if it has the same scope as Cg.
That said, it is possible that the wording is just spin...anyone have more than the supplied link to go off of?
That said, it is possible that the wording is just spin...anyone have more than the supplied link to go off of?
David G:
You are right the choice is OpenGL 2.0 as well as Cg or RenderMonkey (we don't know if it actually IS the same as Cg but we may as well assume it is).
However to 'save face' or to show that ATi is matching/exceeding anything NVIDIA are up to ATi needs something to directly compete with Cg, more for PR and DR and erm TR even (PR we all know, Developer Relations and us guys - Techy Relations).
Whether RenderMonkey will actually get used by developers though is a not that important. That ATi are able to compete in any market that NVIDIA opens up is important for the long run no matter how short sighted it seems.
Integrated motherboard chipsets/platforms
Consoles
Graphic Cards
... these are just some of the markets that ATi and NVIDIA directly compete in so may as well add:
GPU (DX and GL )Compiler Languages
to the list.
RenderMonkey and Cg may actually end up being useful for developers wanting to optimise their engines for the biggest two graphic card designers in the market right now. Or then again maybe not and OpenGL 2.0 will make these solutions a temporary fix in place of a more long term solution (real problem= Assembly Vertex and Shader code is a pain to write).
Edit: mixed up Man and Monkey..... hehe
The name RenderMonkey at least sounds like a slight jab at the seriousness of Cg at least
You are right the choice is OpenGL 2.0 as well as Cg or RenderMonkey (we don't know if it actually IS the same as Cg but we may as well assume it is).
However to 'save face' or to show that ATi is matching/exceeding anything NVIDIA are up to ATi needs something to directly compete with Cg, more for PR and DR and erm TR even
(PR we all know, Developer Relations and us guys - Techy Relations).Whether RenderMonkey will actually get used by developers though is a not that important. That ATi are able to compete in any market that NVIDIA opens up is important for the long run no matter how short sighted it seems.
Integrated motherboard chipsets/platforms
Consoles
Graphic Cards
... these are just some of the markets that ATi and NVIDIA directly compete in so may as well add:
GPU (DX and GL )Compiler Languages
to the list.
RenderMonkey and Cg may actually end up being useful for developers wanting to optimise their engines for the biggest two graphic card designers in the market right now. Or then again maybe not and OpenGL 2.0 will make these solutions a temporary fix in place of a more long term solution (real problem= Assembly Vertex and Shader code is a pain to write).
Edit: mixed up Man and Monkey..... hehe
The name RenderMonkey at least sounds like a slight jab at the seriousness of Cg at least
Going by names alone, RenderMonkey sound like it would be a primitive (think evolution not graphics ) version of RenderMan. So then, yes, this would be ATI's 'Cg'
) version of RenderMan. So then, yes, this would be ATI's 'Cg'Doomtrooper
Veteran
David G. said:
3Dlabs can't lose ground in DX9 or 50% of released titles, so I assume they would enter the 'C' Compiler battleground....'Let's get ready to rumble'
Actualy ... the slowest P10 card with beta drivers get's around 8500 3DMarks in 2001SE so ....
I mean , usualy Pro cards are not so DX optimised but this P10 series seems to be just as goos as the other comercial cards like R8500 and GF4600 .
I think that , with a little driver tweaking , that score could reach up to 9300 - 9500 3DMarks and that the other 2 cards , the ones more powerfull , would reach 11000 3DMarks .
Think just that the fastest card from the P10 series is 50% more powerfull than this one , the one tested by Amazon .
I mean , usualy Pro cards are not so DX optimised but this P10 series seems to be just as goos as the other comercial cards like R8500 and GF4600 .
I think that , with a little driver tweaking , that score could reach up to 9300 - 9500 3DMarks and that the other 2 cards , the ones more powerfull , would reach 11000 3DMarks .
Think just that the fastest card from the P10 series is 50% more powerfull than this one , the one tested by Amazon .
I think Colourless may have hit the nail on the head with the naming.
Assuming this is ATi’s own take on a shader language then I wonder exactly how serious they are with it. The name could suggest not too serious, in fact it could even be a slight dig at Cg. It may be that they intend to use this just to muddy the waters and steer developers away from vendor languages towards DX9 HLSL and OpenGL2. So, I’d like to be in on that meeting to hear what they say.
If they are serious about it then it’s not really going to help matters much, and frankly I’d wonder how far they’d get with it.
Assuming this is ATi’s own take on a shader language then I wonder exactly how serious they are with it. The name could suggest not too serious, in fact it could even be a slight dig at Cg. It may be that they intend to use this just to muddy the waters and steer developers away from vendor languages towards DX9 HLSL and OpenGL2. So, I’d like to be in on that meeting to hear what they say.
If they are serious about it then it’s not really going to help matters much, and frankly I’d wonder how far they’d get with it.
