FX to be canned?

[tart666]

here

Will it be a "limited edition" model?

 

[NFS4]

I call BS

 

[GTaudiophile]

The title of the article is misleading. nVidia will try to sell all 100,000 that they want to produce, and in the mean time, concentrate on NV31, 34, and 35.

 

[Mem]

nVidia will try to sell all 100,000 that they want to produce, and in the mean time, concentrate on NV31, 34, and 35.

That makes sense to me,I don`t think the FX(NV30) will have a very big demand.

 

[chizow]

Not a big surprise. As every fanATic here has stated ad nauseum, the FX is 6 months late. Do you really think the company that invented the 6 month GPU cycle would be waiting around for TSMC to start cranking out the FX in acceptable yields? Now that the .13 micron growing pains have been put to rest, its gonna be good times over at nVidia again, to the tune of something faster and better every six months.

Chiz

 

[Ilmater]

Originally posted by: chizow
Not a big surprise. As every fanATic here has stated ad nauseum, the FX is 6 months late. Do you really think the company that invented the 6 month GPU cycle would be waiting around for TSMC to start cranking out the FX in acceptable yields? Now that the .13 micron growing pains have been put to rest, its gonna be good times over at nVidia again, to the tune of something faster and better every six months.

Chiz

Ditto again. I hate following your posts.

 

[apoppin]

According to the rumour, posted at x-3DFX, Nvidia is canning the GeForce FX and has told its foundry, Taiwan Semiconductor Manufacturing Corp (TSMC), to stop when 100,000 NV30 are produced.

Ok let's get to the SOURCE of the RUMOUR:
A Messagre Board Rumour

But Nvidia UK said this morning it did not comment on speculation posted on bulletin boards.

This is RIDICULOUS! Even for the most fanATIc fanboyz. Who is 3dfx Jedi?

EDIT:

Originally posted by: Ilmater

Originally posted by: chizow
Not a big surprise. As every fanATic here has stated ad nauseum, the FX is 6 months late. Do you really think the company that invented the 6 month GPU cycle would be waiting around for TSMC to start cranking out the FX in acceptable yields? Now that the .13 micron growing pains have been put to rest, its gonna be good times over at nVidia again, to the tune of something faster and better every six months.

Chiz

Ditto again. I hate following your posts.

Don't you have some 3DMark2001 benchmarks to work on "proving" that 1GB RAM is useless?

And IF (somehow) the (ridiculous) rumor is correct - Its "BAD times over at nVidia again".

 

[bluemax]

Originally posted by: chizow
Not a big surprise. As every fanATic here has stated ad nauseum, the FX is 6 months late. Do you really think the company that invented the 6 month GPU cycle would be waiting around for TSMC to start cranking out the FX in acceptable yields? Now that the .13 micron growing pains have been put to rest, its gonna be good times over at nVidia again, to the tune of something faster and better every six months. Chiz

Speaking of "ad nauseum"..... I was beginning to miss you Chiz. Someone had to come in and cheer for nVidia for a while.
I find it hard to believe you own a Radeon 9700 seeing as how you bash ATI and blow trumpets for nVidia wherever possible.
It'd an odd conundrum....

 

[bluemax]

And, for the record, I think this "canning" is a totally unsubstatiated rumour. They wouldn't put this much effort into a product just to can it over a fan issue.
If anything they'd release it with a lower clock, lower price, blow out the inventory and use the knowledge and experience gained with this product to release something better (no a simple refresh) in 6 months or so.
But that's purely speculation as well.

 

[chizow]

LoL Ilmater

Apoppin, I've seen it on 2 different "news" sources, including the FanATic's PC Geek Bible, aka The Inquirer.

nVidia's not gonna waste resources on a product that may only marginally outperform the competition, especially when the $$$ is in the value edition parts (NV31 and NV34). Even though the FX was just announced, its gotta be old news over at nVidia as its specs were completed nearly a year ago. Looks like NV35 is closer than expected...whee!!!!!

Chiz

 

[apoppin]

Originally posted by: chizow
LoL Ilmater

Apoppin, I've seen it on 2 different "news" sources, including the FanATic's PC Geek Bible, aka The Inquirer.

nVidia's not gonna waste resources on a product that may only marginally outperform the competition, especially when the $$$ is in the value edition parts (NV31 and NV34). Even though the FX was just announced, its gotta be old news over at nVidia as its specs were completed nearly a year ago. Looks like NV35 is closer than expected...whee!!!!!

Chiz

Well it's not at Rage3D so it mustn't be true . . .

It's a RUMOUR . . . since you "believe" it, it speaks volumes about your own "credibility". And again - IF it is TRUE - nVidia is in huge TROUBLE.

:Q

 

[chizow]

Originally posted by: bluemax
Speaking of "ad nauseum"..... I was beginning to miss you Chiz. Someone had to come in and cheer for nVidia for a while.
I find it hard to believe you own a Radeon 9700 seeing as how you bash ATI and blow trumpets for nVidia wherever possible.
It'd an odd conundrum....

Its a Love-Hate relationship. I love the hardware; I hate the software

And I don't blow trumpets for them, I just bash the misinformation

And, for the record, I think this "canning" is a totally unsubstatiated rumour. They wouldn't put this much effort into a product just to can it over a fan issue.
If anything they'd release it with a lower clock, lower price, blow out the inventory and use the knowledge and experience gained with this product to release something better (no a simple refresh) in 6 months or so.
But that's purely speculation as well.

I expected something like this from nVidia, I just didn't think the .13 process would be mature enough to support full production of NV35 this soon. If you REALLY think about it, this makes perfect sense. Its clear that right now, its still difficult for TSMC to obtain high yields on NV30 at the necessary clock speeds for the FX. What does that leave you with? A much cheaper, cooler, slower, less power consuming chip for your value and mobile lines. After all, thats all the value parts are anyways; simplified or slower cores based on the flagship model. Its not simply a fan issue, the FX is gonna be thrown out as a stop-gap to launch their .13 product lines, and followed soon after by NV35. There's simply no reason for them to hold back superior technology for the sake of sustaining "old" tech as others have stated

Chiz

 

[Mem]

I remember reading Nvida have two main engineering teams that work on different card products,so it`s possible that the NV35 could be nearer then we think if they are still using the same system.

 

[bluemax]

I expected something like this from nVidia, I just didn't think the .13 process would be mature enough to support full production of NV35 this soon. If you REALLY think about it, this makes perfect sense. Its clear that right now, its still difficult for TSMC to obtain high yields on NV30 at the necessary clock speeds for the FX. What does that leave you with? A much cheaper, cooler, slower, less power consuming chip for your value and mobile lines. After all, thats all the value parts are anyways; simplified or slower cores based on the flagship model. Its not simply a fan issue, the FX is gonna be thrown out as a stop-gap to launch their .13 product lines, and followed soon after by NV35. There's simply no reason for them to hold back superior technology for the sake of sustaining "old" tech as others have stated Chiz

Great googaly-moogaly! We just agreed on something! :Q :Q :Q

 

[Lint21]

If this is true, it would make a lot of sense. Everything about the FX so far has been, to put it lightly, lame. Nvidia has always proven to be a smart company; they didn't get to where they are (#1, like it or not) by making bad business decisions. They've got to see what a joke their product is. I can't imagine they wouldn't act on that.

The logical thing for them to do is to pull the plug before this thing takes on a life of it's own. They'll try to make it low-key, like it was planned, but pull the plug nonetheless. At least that's whay I would do; quietly acknowledge defeat, then come back ASAP with a better product.

 

[Lonyo]

Originally posted by: Mem
I remember reading Nvida have two main engineering teams that work on different card products,so it`s possible that the NV35 could be nearer then we think if they are still using the same system.

I thought the $400 million R&D was for the NV30 and NV35 lines, from NV30 to the NV39, which would mean a lot of the R&D on the 35 has already been done and they're possibly edging up on doing some NV35 finalisations.
If the 2 lines are very similar, like the GeForce 3 to 4 shift, then a 6 month cycle should be back in the picture, but possibly not a major performance boost with the next itineration unless yeild and drivers start really getting somewhere.

 

[chizow]

Originally posted by: Lonyo

Originally posted by: Mem
I remember reading Nvida have two main engineering teams that work on different card products,so it`s possible that the NV35 could be nearer then we think if they are still using the same system.

I thought the $400 million R&D was for the NV30 and NV35 lines, from NV30 to the NV39, which would mean a lot of the R&D on the 35 has already been done and they're possibly edging up on doing some NV35 finalisations.
If the 2 lines are very similar, like the GeForce 3 to 4 shift, then a 6 month cycle should be back in the picture, but possibly not a major performance boost with the next itineration unless yeild and drivers start really getting somewhere.

Well, the hard part (designing the core and fab'ing it on a .13 process with TSMC is completed), the obstacles that remain are improving the fab process for a higher yielding, cooler, less power-hungry core so that the chip can scale. All have been accomplished through core revisions and fab improvements historically, I don't think the situation here will be any different. The easy part will be mating a slightly revised core (which looks to be killer from prelim results) with a 256-bit memory interface which can truly take advantage of DDR-II. Enter NV35

Chiz

 

[Pocatello]

100,000 GFFX is a good number for nvidia.

 

[codehack2]

Originally posted by: chizow

Originally posted by: Lonyo

Originally posted by: Mem
I remember reading Nvida have two main engineering teams that work on different card products,so it`s possible that the NV35 could be nearer then we think if they are still using the same system.

I thought the $400 million R&D was for the NV30 and NV35 lines, from NV30 to the NV39, which would mean a lot of the R&D on the 35 has already been done and they're possibly edging up on doing some NV35 finalisations.
If the 2 lines are very similar, like the GeForce 3 to 4 shift, then a 6 month cycle should be back in the picture, but possibly not a major performance boost with the next itineration unless yeild and drivers start really getting somewhere.

Well, the hard part (designing the core and fab'ing it on a .13 process with TSMC is completed), the obstacles that remain are improving the fab process for a higher yielding, cooler, less power-hungry core so that the chip can scale. All have been accomplished through core revisions and fab improvements historically, I don't think the situation here will be any different. The easy part will be mating a slightly revised core (which looks to be killer from prelim results) with a 256-bit memory interface which can truly take advantage of DDR-II. Enter NV35

Chiz

I agree, and not to sound like I'm bashing Nvidia, but based on thier past products, the X5 release seems to be the one that get's it right. Get the technology out the door with the 1st release and then perfect it with the second. Looks like NV30 to NV35 will hold true to this model.

CH2

 

[dullard]

Didn't all the previews all state that the NV30 will be a limited edition. NVidia was supposedly using it to get a foot in the 0.13 micron door and they are more focused on the NV31+ cards. I'm pretty sure I've read this many times over the last several months. So no surprize here.

 

[GoodRevrnd]

Originally posted by: dullard
Didn't all the previews all state that the NV30 will be a limited edition. NVidia was supposedly using it to get a foot in the 0.13 micron door and they are more focused on the NV31+ cards. I'm pretty sure I've read this many times over the last several months. So no surprize here.

 

[mchammer187]

Originally posted by: chizow

Originally posted by: Lonyo

Originally posted by: Mem
I remember reading Nvida have two main engineering teams that work on different card products,so it`s possible that the NV35 could be nearer then we think if they are still using the same system.

I thought the $400 million R&D was for the NV30 and NV35 lines, from NV30 to the NV39, which would mean a lot of the R&D on the 35 has already been done and they're possibly edging up on doing some NV35 finalisations.
If the 2 lines are very similar, like the GeForce 3 to 4 shift, then a 6 month cycle should be back in the picture, but possibly not a major performance boost with the next itineration unless yeild and drivers start really getting somewhere.

Well, the hard part (designing the core and fab'ing it on a .13 process with TSMC is completed), the obstacles that remain are improving the fab process for a higher yielding, cooler, less power-hungry core so that the chip can scale. All have been accomplished through core revisions and fab improvements historically, I don't think the situation here will be any different. The easy part will be mating a slightly revised core (which looks to be killer from prelim results) with a 256-bit memory interface which can truly take advantage of DDR-II. Enter NV35

Chiz

maybe GTAudiophile was right than

Geforce FX 5800 Ultra is DOA?

i am certainly looking forward to an early NV35 if this is true though

 

[Adul]

Originally posted by: mchammer187

Originally posted by: chizow

Originally posted by: Lonyo

Originally posted by: Mem
I remember reading Nvida have two main engineering teams that work on different card products,so it`s possible that the NV35 could be nearer then we think if they are still using the same system.

I thought the $400 million R&D was for the NV30 and NV35 lines, from NV30 to the NV39, which would mean a lot of the R&D on the 35 has already been done and they're possibly edging up on doing some NV35 finalisations.
If the 2 lines are very similar, like the GeForce 3 to 4 shift, then a 6 month cycle should be back in the picture, but possibly not a major performance boost with the next itineration unless yeild and drivers start really getting somewhere.

Well, the hard part (designing the core and fab'ing it on a .13 process with TSMC is completed), the obstacles that remain are improving the fab process for a higher yielding, cooler, less power-hungry core so that the chip can scale. All have been accomplished through core revisions and fab improvements historically, I don't think the situation here will be any different. The easy part will be mating a slightly revised core (which looks to be killer from prelim results) with a 256-bit memory interface which can truly take advantage of DDR-II. Enter NV35

Chiz

maybe GTAudiophile was right than

Geforce FX 5800 Ultra is DOA?

i am certainly looking forward to an early NV35 if this is true though

so are u saying in a way Chiz agrees with Gtaudiofile?




I think nvidia will produce some, i can see it has a collectors item

 

[mchammer187]

Originally posted by: Adul

Originally posted by: mchammer187

Originally posted by: chizow

Originally posted by: Lonyo

Originally posted by: Mem
I remember reading Nvida have two main engineering teams that work on different card products,so it`s possible that the NV35 could be nearer then we think if they are still using the same system.

I thought the $400 million R&D was for the NV30 and NV35 lines, from NV30 to the NV39, which would mean a lot of the R&D on the 35 has already been done and they're possibly edging up on doing some NV35 finalisations.
If the 2 lines are very similar, like the GeForce 3 to 4 shift, then a 6 month cycle should be back in the picture, but possibly not a major performance boost with the next itineration unless yeild and drivers start really getting somewhere.

Well, the hard part (designing the core and fab'ing it on a .13 process with TSMC is completed), the obstacles that remain are improving the fab process for a higher yielding, cooler, less power-hungry core so that the chip can scale. All have been accomplished through core revisions and fab improvements historically, I don't think the situation here will be any different. The easy part will be mating a slightly revised core (which looks to be killer from prelim results) with a 256-bit memory interface which can truly take advantage of DDR-II. Enter NV35

Chiz

maybe GTAudiophile was right than

Geforce FX 5800 Ultra is DOA?

i am certainly looking forward to an early NV35 if this is true though

so are u saying in a way Chiz agrees with Gtaudiofile?




I think nvidia will produce some, i can see it has a collectors item

as amazing as it seems yup

producing only 100,000 = DOA in my book and most likely chiz's

 

[paralazarguer]

FX to be canned?

Yeah, that would be a really great way to ship them. Retailers could tell if the can had been opened and it helps keep the freshness in.