View Full Version : NV30 status, rally not taped out? + more than NV30 in Q4
Hey everyone,
Before anyone say " Did you even listen to that conference call? " - yes. I heard all of it live.
Anyway, there are 2 explanations to this evasion:
1. NV30 didn't tape-out yet.
2. NV30 taped out, but VERY recently. Jen Hsun, not wanting to make stupid analysts barely knowing what a GPU is, said it will be out in the holiday season. It could have taped out two days before - but since he said it was nearly taped out in the last conference call... he probably would hate to give them an info being able to proof he lied.
Both 1 and 2 are possible, but 2 was not said by anyone, and i find it nearly as likely as 1 ( but still not as likely ) - just wanted to give a possibility no one talked about.
Also, it sounds like we ARE going to have multiple GPUs in Q4 ( Jen Hsun said it ), not only the NV30. As to how much GPUs, i got to refer to a Prudential analyst, which, a few months ago, said 9 chips will be available. That's less than the GF4, but some parts of the GF4 family do not need as much to be replaced - nVidia still pretty much got the lead in workstation, they'll only need a new higher end i guess.
Uttar
netviper13
08-16-02, 04:46 PM
A couple recent articles confirmed that the final version of NV30 taped out, and should be ready to ship within the usual time from tapeout to ship.
intercede007
08-16-02, 05:09 PM
Yeah, anandtech has an article that says the NV30 taped-out only within the past week or two. I've also heard rumors that there was a beta-board of the NV30 using the .15 micron process a couple months ago. Anyone hear that to and/or have an information about that?
Who cares when nv30 tapes out?If the ceo said that nv30 will ship in december it will ship then!!Well he has also said before that nv30 would be the first DX9 card and that it would ship in august!!:rolleyes:
netviper13
08-16-02, 07:28 PM
Ideally it would have been the first DX9 card out, but they had several things go wrong.
First is they over-stretched themselves, with working on X-Box, nForce, and GeForce4 at the same time. Then they had problems with the .13 process (as nearly everybody who makes the switch does at first).
Bigus Dickus
08-16-02, 07:34 PM
Yes, but somehow designing an IGP, GPU, and game console chip didn't set back ATI quite so far. ;)
Must be that pesky .13u... oh wait, taping out isn't TSMC's responsibility is it. :D
netviper13
08-16-02, 07:37 PM
ATI fans seem to really harp on nVidia for NV30 delays, but it's comparing apples to oranges.
Radeon 9700 is the evolutionary step of the 8500, whereas NV30 is truly revolutionary.
It takes time to do what they did with NV30, mainly because it had not been done before!
StealthHawk
08-16-02, 10:50 PM
Originally posted by Bigus Dickus
Yes, but somehow designing an IGP, GPU, and game console chip didn't set back ATI quite so far. ;)
Must be that pesky .13u... oh wait, taping out isn't TSMC's responsibility is it. :D
because ATI didn't design the game console chip :rolleyes:
ATI's motherbaords have been far from perfect too, are they even out yet? i remember seeing reviews or previews months ago and hearing about poor integrated graphics performance.
jbirney
08-16-02, 11:12 PM
Radeon 9700 is the evolutionary step of the 8500, whereas NV30 is truly revolutionary.
OMFG, tell me your joking right? You forget that there are only a few things that set the NV30 and R9700 a part. The NV has can excuite more vertex/pixel shaders per pass and has better flow control. Thats it. Also the R9700 can do everything that is publicy know about the NV30 it just takes more pases. This is just like the PS1.4 on the R8500 vrs PS 1.3 on the GF4. Please explain how two parts with almost the same feature set can be so different?
It takes time to do what they did with NV30, mainly because it had not been done before!
The only really new thing the nV has is its made on the .13 micron process which is not revolutionary.
Bigus Dickus
08-17-02, 12:48 AM
Originally posted by StealthHawk
because ATI didn't design the game console chip :rolleyes: ATI has publicly stated that the former ArtX team (don't remember if they call that one east or west) was prime on the R300 development. :rolleyes:
So yeah, they did design a kick ass console chip and a kick ass PC graphics chip, on time.
There are plenty of legitimate reasons for the NV30 to be late, but if "being stretched too thin" is one of them then NVIDIA simply made a mistake, because it's not impossible to do.
StealthHawk
08-17-02, 02:31 AM
ATI bought ArtX. they were already working on Flipper. that's different from say reassigning your engineers to work on a new chip instead of working on one they would have been previously working on. AFAIK ATI didn't move anyone over to the ArtX team, more it was the other way around(ArtX joining ATI guys for the R300).
Bigus Dickus
08-17-02, 02:43 AM
ArtX designed the Flipper, and that group became a separate design team that then designed the R300. I don't see how that is in any way different, since the same group of guys did the work on both projects. ArtX, ATI, call that group of engineers what you want, but the bottom line is that it didn't delay them, and it did NVIDIA's engineers.
netviper13
08-17-02, 03:20 AM
Because nVidia didn't ACQUIRE the chip, they designed it from the ground up. The ART-X team had already nearly finished the Flipper chip, and about the only thing ATI had to do with it is that it now gets branded under the ATI name.
XGPU, however, was designed from the ground up by nVidia's engineers.
Bigus Dickus
08-17-02, 03:40 AM
sigh.
I don't care what name you give to them. One group of engineers designed the Flipper chip, and then turned around and designed the R300. The same guys. Both chips.
If one group of engineers could do it, why couldn't NVIDIA's engineers? :rolleyes:
StealthHawk
08-17-02, 06:28 AM
ArtX engineers designed R300 entirely on their own without any input whatsoever from other engineers working at ATI? unless that above statement is true, the parallel you are drawing is not accurate.
SurfMonkey
08-17-02, 09:57 AM
I guess the distinction that people are making is that ATi bought ready made technology and shoe horned it into a PC chip. nVidia, on the other hand, bought the rights to a whole load of theoretical technologies and had to do their design from the ground up.
An interesting aside, if JC has built the NV30 backend to DoomIII then what did he use as a reference when he did the programming? Has he now got a version of the NV30 to play with?
Originally posted by SurfMonkey
I guess the distinction that people are making is that ATi bought ready made technology and shoe horned it into a PC chip. nVidia, on the other hand, bought the rights to a whole load of theoretical technologies and had to do their design from the ground up
I don't get it, what load of theoretical technologies has nv30 over radeon 9700?Do you know something we don't?And don't tell me about the cinefx, please!!!:rolleyes:
If he has a board then why not use it for quakecon presentation?
Anyway it's good to hear that he might have an nv30 because it might be closer to launch!!At least I hope!!
SavagePaladin
08-17-02, 10:34 AM
because the 9700 drivers are close to finished :rolleyes:
Originally posted by SavagePaladin
because the 9700 drivers are close to finished :rolleyes:
If nv30 is so powerful and 2x faster than radeon 9700(as many people here have been saying) it should beat it hands down even with alpha drivers!!:rolleyes:
SavagePaladin
08-17-02, 11:11 AM
Beating and working right don't always go together.
Originally posted by Fotis
If nv30 is so powerful and 2x faster than radeon 9700(as many people here have been saying) it should beat it hands down even with alpha drivers!!:rolleyes:
NV30 will be significantly faster than R9700, but let's stay realistic here, nobody really expects it to double the perfomance of R9700! :)
SurfMonkey
08-17-02, 12:01 PM
Originally posted by Fotis
I don't get it, what load of theoretical technologies has nv30 over radeon 9700?Do you know something we don't?And don't tell me about the cinefx, please!!!:rolleyes:
If he has a board then why not use it for quakecon presentation?
Anyway it's good to hear that he might have an nv30 because it might be closer to launch!!At least I hope!!
I never said that the NV30 has anything over and above the R300 technologie wise, though that's not strictly true; .13u process being one example. But it is fairly obvious that it will have a higher core and memory speed than the R300.
What I meant was the Gigapixel and 3dfx technologies were all theoretical, nothing had ever been, successfully, built with them. nVidia had to do all the design work and implementation work from scratch. ATI bought a ready made design and re-engineered to become a PC gfx chip.
If he has a board then he will be under a very strict NDA and, unless nVidia allow him to demo the NV30, he won't be able to use it publically.
Originally posted by SurfMonkey
I never said that the NV30 has anything over and above the R300 technologie wise, though that's not strictly true; .13u process being one example. But it is fairly obvious that it will have a higher core and memory speed than the R300.
What I meant was the Gigapixel and 3dfx technologies were all theoretical, nothing had ever been, successfully, built with them. nVidia had to do all the design work and implementation work from scratch. ATI bought a ready made design and re-engineered to become a PC gfx chip.
It'll be a surprise for me if nv30 is clocked higher than radeon 9700,remember gf2ultra=250 & gf3=200!!On the other hand they have some new blood(3dfx)things may be different this time.
I don't expect nvidia to use any of the rampage-mojo tech in nv30 because if I remember corectly David Kirk said that they mixed the 3dfx engineers with nvidias to make up new tech.
I don't think that radeon was a re-engineered design.
SavagePaladin
08-17-02, 04:20 PM
The memory speed at the very least is known to be higher. Thats one rumor I don't doubt.
Anyway, carry on while I go back to sleep.
Originally posted by SavagePaladin
The memory speed at the very least is known to be higher. Thats one rumor I don't doubt.
Anyway, carry on while I go back to sleep.
Did you drink your milk first?
I know the memory will be higher I was talking about core clock.
Everything I right is based on facts and past performance by nvidia.;)
vBulletin® v3.7.1, Copyright ©2000-2009, Jelsoft Enterprises Ltd.