PDA

View Full Version : NV36 to be manufactured at IBM


Uttar
06-26-03, 07:36 AM
Hey,


I'm lazy, so I'll just quote GPU:RW (http://www.notforidiots.com/GPURW.php) ( read: myself ) ;)

In light of new information, I suggest everyone not to trust any of the NV36 info we had so far. A lot of it is very wrong, and that means the rest most likely is. First of all, the NV36 will be manufactured at IBM's 0.13µ process. Secondly, the NV36 will not fully replace the NV31 ( at least, not the flipchip version ). And finally, it's the NV36 which taped-out at IBM, not the NV40. The NV40 didn't tape-out yet. What market segment is the NV36 in, then? I don't know. Most likely $199-$299, with th NV31 Regular replacing the NV34 Ultra, and the Ultra being between the NV36 Regular and the NV31 Regular. No idea about that part though, it's just speculation, but those three "facts" are very reliable.

There, now that'll teach me to assume the NV36 was done at TSMC because it's more of a NV31 replacement :P Bah!


Uttar

Paul
06-26-03, 07:47 AM
Interesting information, which provides some answers for the points I raised in another thread, but which were never answered:

Originally posted by Paul
Just listened to the conference call, and i've got two points:

1) As far as I can tell, he never mentions anything taping out. Unless i missed it during a couple of places where the sound dies out a little, this seems to have been misreported.

2) He doesn't sound that enthusiastic about the new product for the fall. He talks about it simply supplimenting the current family, so that if it doesn't turn up it's no big deal (which doesn't sound like the NV40 - more like the NV36, but as Uttar already said, that doesn't make sense). He then contradicts himself a little later, by saying nV are replacing the entire family in the fall.

Can anyone clear this up?

SurfMonkey
06-26-03, 10:13 AM
So I was right that the NV36 had taped out at IBM? Cool, and here I was doubting my sources. :p

Slappi
06-26-03, 10:43 AM
Originally posted by Paul
Interesting information, which provides some answers for the points I raised in another thread, but which were never answered:

What if they just skipped the whole NV36-39 and released just the NV 40,41,42,43,44 line this fall?

Is that possible?

GlowStick
06-26-03, 03:46 PM
Wow, thats very cool, i didnt know the nv36 was gonna be made at ibm, but i wonder if that dose mean 'insane' clock speeds!

i bet the nv36 wont have an ungodly amount of transistors, and ibm 'should' beable to make a better chip!

woudlnt it be cool if we got up to 700mhz clock speeds on it : P

SHS`
06-26-03, 04:05 PM
Originally posted by GlowStick
Wow, thats very cool, i didnt know the nv36 was gonna be made at ibm, but i wonder if that dose mean 'insane' clock speeds!

i bet the nv36 wont have an ungodly amount of transistors, and ibm 'should' beable to make a better chip!

woudlnt it be cool if we got up to 700mhz clock speeds on it : P

I'd guess this is nVidia/IBM in testing the process in production... much easier and less risky to run it on lowend stuff, but there's a good chance they would overclocxk well, but memory would be the limiting factor as they wouldn't use the highest grade stuff.

GlowStick
06-26-03, 04:10 PM
If the nv40 didnt tape out, then i feel like its pritty late for it!

worries me.

Paul
06-26-03, 04:15 PM
It wasn't due to tape out until next month anyway, so it was already early at best.

SurfMonkey
06-26-03, 07:52 PM
I get the feeling that NVidia looked upon the 9600Pro and saw that it was good, and in doing so decided to do the same as ATI did. And that is to use a smaller, less complex, chip to get used to a design process and to work out the best recipe to use at the foundry.

BigFish7
06-28-03, 12:25 PM
Originally posted by SurfMonkey
I get the feeling that NVidia looked upon the 9600Pro and saw that it was good, and in doing so decided to do the same as ATI did. And that is to use a smaller, less complex, chip to get used to a design process and to work out the best recipe to use at the foundry.

good point there :) i think they should have done it earler... and then run a respin for the high end parts...
yet i think its NV3X bad silicon design, lets remember they did just fine with the Ti line...
hmm they just tried to hard :) i hope the NV40 would put end to this doubt... :D