Is it set in stone..

[overclocked]

that R420 don´t have PS3? Been away for a couple of months and just looked at the NV40 Reviews i was shocked, surprised etz.. What happend too the leaked specs of 180 transistors and build "around" nV3x series?

Also one thing i thought of regarding NV45 is the manufactoring process it will be build on and who will make it, IBM or the "other" ...
Many questions and more too come, but please speculate.

 

[Evildeus]

According to The Baron, the NV45 is on 0.13 low-k already tape out and from TSMC. As for R420 and PS3.0 it's believed that there won't be PS3.0, but time will tell.

 

[overclocked]

No 0.11micron shrink for NV45 then.

 

[Joe DeFuria]

No 0.11micron shrink for NV45 then.

I have doubts that that ANY of the high-end cards will be produced on 0.11. More likely, IMO, the mid and low-end parts will be migrated to 0.11, and the high-end replacements will go directly to 0.09. Just a guess.

 

[Boke]

I'm wondering if it will support PS2.x.

 

[tazdevl]

What else would it support?

 

[tb]

2_b -> 2.0 with max. 512 instructions instead of the 96 from 2_0

 

[Boke]

PS2.x is not the same as PS2.0 . Its much closer to 3.0 than it is 2.0 .

 

[tEd]

PS2.x is not the same as PS2.0 . Its much closer to 3.0 than it is 2.0 .

what tb said which makes r420 already ps2.x

 

[Boke]

I thought 2.x was 96 minimum with the upper limit specified in the PS caps. I'll have to look at the sdk later.

 

[tb]

PS2.x is not the same as PS2.0 . Its much closer to 3.0 than it is 2.0 .

- ps_2_x without additional caps is just ps_2_0 (that's why ps_2_x and ps_2_0 shaders run on the r3xx)

- ps_2_x with all additional caps is very close to ps_3_0

Thomas

 

[tEd]

PS2.x is not the same as PS2.0 . Its much closer to 3.0 than it is 2.0 .

- ps_2_x without additional caps is just ps_2_0 (that's why ps_2_x and ps_2_0 shaders run on the r3xx)

- ps_2_x with all additional caps is very close to ps_3_0

Thomas

for what kind of featureset does the ps2.x hlsl target compile to exactly?

 

[nelg]

[lazy]

Bear with my ignorance here guys but looking at this it seems that there is not to much of a difference WRT PS2.x vs PS3.0. Even with the loop/endloop instructions could not just use rep/endrep instead with an arbitrarily high counter?

ps_3_0 Features
New features:

Consolidated 10 Input Registers (v#)
Indexable Constant Float Register (c#) with Loop Counter Register (aL)
Number of Temporary Registers (r#) increased to 32
Number of Constant Float Registers (c#) increased to 224
New instructions:

Setup instruction - dcl_usage
Static flow instructions - loop, endloop
Arithmetic instruction - sincos (new syntax)
Texture instruction - texldl
New registers:

Input Register (v#)
Position Register (vPos)
Face Register (vFace)

[/lazy]

 

[Ostsol]

for what kind of featureset does the ps2.x hlsl target compile to exactly?

Pixel Shader Differences

 

[Megadrive1988]

by the time Pixel Shader 3.0 / 3.x *really* matters, ATI will have the r500 out. probably summer or fall of next year. at that time, Nvidia might be struggling to get the NV50 out.

 

[trinibwoy]

by the time Pixel Shader 3.0 / 3.x *really* matters, ATI will have the r500 out. probably summer or fall of next year. at that time, Nvidia might be struggling to get the NV50 out.

Bias reading :

Nvidia[--------|-------X]Ati

 

[Dave Baumann]

Why would they "struggle"? (And why wouldn't ATI with R500?)

 

[SmuvMoney]

After reading the differences in the link above (thanks Ostsol), I have one question: if ATi fully supports ps_2_x to the letter, will nVidia's support of ps_3_0/SM3.0 still be a big deal as some are (IMHO mistakenly) making it? After reading the differences, I don't think so, but I don't claim to be in the know. I also do not claim to fully understand how significant the changes are from ps_2_x to ps_3_0 especially with regard to the extra registers needed to support it. It looks like the jump from ps_2_0 to ps_2_x was more significant in terms of added functionality where the jump from ps_2_x to ps_3_0 looks like it is just making it faster and/or easier. Am I missing something blatantly obvious like the noob I am?

 

[Ardrid]

Why would they "struggle"? (And why wouldn't ATI with R500?)

Dave, check your PM plz

 

[trinibwoy]

After reading the differences in the link above (thanks Ostsol), I have one question: if ATi fully supports ps_2_x to the letter, will nVidia's support of ps_3_0/SM3.0 still be a big deal as some are (IMHO mistakenly) making it? After reading the differences, I don't think so, but I don't claim to be in the know. I also do not claim to fully understand how significant the changes are from ps_2_x to ps_3_0 especially with regard to the extra registers needed to support it. It looks like the jump from ps_2_0 to ps_2_x was more significant in terms of added functionality where the jump from ps_2_x to ps_3_0 looks like it is just making it faster and/or easier. Am I missing something blatantly obvious like the noob I am?

Nope I think you hit the nail on the head first time. If Nvidia can convince developers to implement certain things with good performance in 3.0 that cannot be done with equivalent performance in 2.0 there may be a game on and that's a big IF. But we're in the same noobish boat so who knows