There was a strange post by Dave over at Beyond3D last night that had a whole lot of fun speculating, but just now on page eleven either Baron spilled the beans or he's farking around with the Dig again:
The Baron at Beyond3D
Okay. Time to end the speculation.
Note that you haven't seen a mobile NV40. Now, you're all thinking, "Oh, it's obviously because of the NV40's massive heat output and thermal requirements." WRONG!
Enter NVM1. The first use of Gigapixel IP. It's a mobile-only chip that has absolutely nothing in common with NV40 (or at least, the NV40 you know now--remember that eight-pipe "NV40" making the rounds earlier this year? that wasn't an NV40, that was an NVM1 in handy AGP form for developers to play with.).
The whole point of it is to be competitive with NV40U in a package small enough and cool enough for a mobile chip. This is not going to be a Mobility 9700 or something like that--you are not going to find it on sub-$3000 notebooks. This is a chip for the truly insane. Last I heard about it, it was an 8x1, but that was several months ago, and it has apparently changed significantly since then.
It taped out at UMC because of IBM's fabbing problems and TSMC's capacity problems. At this point, I don't know much else about it. AA? Dunno. It does support PS/VS3.0, and it's a lot closer to the level of Series 5 and all that then NV40 is.
Either way, though, it's fast as hell. I also hear whisperings that it incorporates a lot of design principles from 3dfx's last project, the post-Rampage unnamed chip that was to come out around this generation.
You can check out Baron's post at Beyond3D and see how it develops or read the thread at Elite Bastards where Baron talks a bit more about it too.
Okay. Time to end the speculation.
Note that you haven't seen a mobile NV40. Now, you're all thinking, "Oh, it's obviously because of the NV40's massive heat output and thermal requirements." WRONG!
Enter NVM1. The first use of Gigapixel IP. It's a mobile-only chip that has absolutely nothing in common with NV40 (or at least, the NV40 you know now--remember that eight-pipe "NV40" making the rounds earlier this year? that wasn't an NV40, that was an NVM1 in handy AGP form for developers to play with.).
The whole point of it is to be competitive with NV40U in a package small enough and cool enough for a mobile chip. This is not going to be a Mobility 9700 or something like that--you are not going to find it on sub-$3000 notebooks. This is a chip for the truly insane. Last I heard about it, it was an 8x1, but that was several months ago, and it has apparently changed significantly since then.
It taped out at UMC because of IBM's fabbing problems and TSMC's capacity problems. At this point, I don't know much else about it. AA? Dunno. It does support PS/VS3.0, and it's a lot closer to the level of Series 5 and all that then NV40 is.
Either way, though, it's fast as hell. I also hear whisperings that it incorporates a lot of design principles from 3dfx's last project, the post-Rampage unnamed chip that was to come out around this generation.
You can check out Baron's post at Beyond3D and see how it develops or read the thread at Elite Bastards where Baron talks a bit more about it too.
