According to Dave Orton they had/have 3 development teams. Marborough, Silicon Valley, Orlando. Quote from the interview at B3D with Dave Orton in 2004:
Source: http://beyond3d.com/content/interviews/8/3
nvidia "D8E" High End solution, what can we expect in 2008?
- Thread starter 2900guy
- Start date
According to Dave Orton they had/have 3 development teams. Marborough, Silicon Valley, Orlando. Quote from the interview at B3D with Dave Orton in 2004:
This is actually a good question b/c Fudo is the only one in the industry who's using this codename =) And from the sound of it i'd say that it looks more like a board codename -- a GT-level board with two chips on it...
So do you mean GT200 is no new GPU codename but rather something like DualCore GPU with two G100 GPUs??
But don`t you think it`s something strange? I`m talking about GF9800GTX/GT which are rumoured to be launched in March/April. There are also some rumours that GF9800s will be based on another G92 SKUs (G92-450 G92-470??). Does it make sense releasing GF9800 and 3 or 4 months later another highend GPU ??(G100?? GT200??)
No no no, AFAI_Heard there can't be such thing as a dual G100 board b/c well G100 is REALLY big (in this sence Fudo's right when he says that "GT200" will be even hotter then G92GX2 -- but i think that he made up the "hotter" part from the "really big chip" part by himself which isn't necessarily true).
Nothing is for certain from NV.But don`t you think it`s something strange? I`m talking about GF9800GTX/GT which are rumoured to be launched in March/April. There are also some rumours that GF9800s will be based on another G92 SKUs (G92-450 G92-470??). Does it make sense releasing GF9800 and 3 or 4 months later another highend GPU ??(G100?? GT200??)
I'm not even 100% sure right now that we'll even see a 9800GX2 on two G92s at all.
As for the GT200 and Fudo in general -- i have a feeling that while he does have some kind of insider in ATI/AMD, all of the NV news he's just mading up himself.
I could be wrong though about that and GT200 can really exist but what if it's not a chip but a board for GF9800GX2? Two GTs (G92=8800GT) slapped together... GT200...
I don't know, it's just that i can't seem to find any reference to something called GT200 anywhere but at Fudo's place -- and that makes me question the origin of this thing.
How do you know that G100 is - as you are saying - REALLY big and why dou you think there is no possibility to make DualCore GPU with 2xG100?? Probably AMD will push r700 which will be DualCore rv770 so why NVIDIA couldn`t do the same thing with G100?
Cause you can't do a dual chip board with two G80s, and from what i hear G100 is a bigger chip than G80.How do you know that G100 is - as you are saying - REALLY big and why dou you think there is no possibility to make DualCore GPU with 2xG100?? Probably AMD will push r700 which will be DualCore rv770 so why NVIDIA couldn`t do the same thing with G100?
Silent_Buddha
Legend
This year will certainly be interesting with rumors suggesting Nvidia will continue with the large monolithic chip while ATI is focused on multiple smaller chips.
Unless ATI comes up with something significantly more elegant than AFR, I see Nvidia continuing to dominate the high end, while ATI moves into a dominant position in the mid/low range while battling it out in the performance sector.
Regards,
SB
Unless ATI comes up with something significantly more elegant than AFR, I see Nvidia continuing to dominate the high end, while ATI moves into a dominant position in the mid/low range while battling it out in the performance sector.
Regards,
SB
Yes i know DualG80 like GPU is impossible to do. You are saying G100 is even bigger chip than G80?? Well, it means it is really big chip - about 500mm^2?? I hope performance will be great too. If rv770 is said to be about 50% or more faster than rv670, G100 must be at least 50-70% faster than G80.
That would imply that RV670 is faster than G80, which is clearly not the case (in fact, it's quite the opposite).
No no i mean if G100 is big chip then it will be more expensive than Rv770 so must be faster (i.e. G92 vs rv670 situation) than it
Well if Fuad is right about the RV770 performance then G100 should be more than 2 times faster than RV770. So even dual-chip RV770 board won't be able to compete with it. And if that is true then i think we'll see another one of AMD's "we're slower but cheaper" things. Only this time their _dual-chip_ AFR-board will be slower than one G100 and this is gonna hurt.
But those are rumours after all
They tend to be inaccurate and even false.
And even if they are true today everything can chang in the nearest future.
So you never know... Until the product is released.
But those are rumours after all
They tend to be inaccurate and even false.
And even if they are true today everything can chang in the nearest future.
So you never know... Until the product is released.
Well if Fuad is right about the RV770 performance then G100 should be more than 2 times faster than RV770. So even dual-chip RV770 board won't be able to compete with it. And if that is true then i think we'll see another one of AMD's "we're slower but cheaper" things. Only this time their _dual-chip_ AFR-board will be slower than one G100 and this is gonna hurt.
But those are rumours after all
They tend to be inaccurate and even false.
And even if they are true today everything can chang in the nearest future.
So you never know... Until the product is released.
You've mastered the art of teasing, that much is certain
.Sound_Card
Regular
Well if Fuad is right about the RV770 performance then G100 should be more than 2 times faster than RV770. So even dual-chip RV770 board won't be able to compete with it. And if that is true then i think we'll see another one of AMD's "we're slower but cheaper" things. Only this time their _dual-chip_ AFR-board will be slower than one G100 and this is gonna hurt.
But those are rumours after all
They tend to be inaccurate and even false.
And even if they are true today everything can chang in the nearest future.
So you never know... Until the product is released.
Seems little silly comparing a RV to a flagship aye? Assuming R700 is not a AFR solution and that indeed RV770 is 50%+ faster than RV670, that would imply R700 is roughly 200%(aka nearly three times) faster than R600.
Isn't that what AMD is doing right now, comparing the HD3870X2 against the 8800 GTX/Ultra ?
Sound_Card
Regular
Isn't that what AMD is doing right now, comparing the HD3870X2 against the 8800 GTX/Ultra ?
R680.
And it seems the two chip RV770 has a code name going by R700. See the pattern?
Now R700 being another AFR solution is the million dollar question.
Guess I misread the thread title.
"R680" is the codename for the card, not the GPU's in it...
What?? If Rv770 is about 50% faster than r670 how could you get 200% when r700 will be 2xRv770?? It seems it will be about 2 times faster not 3
Sound_Card
Regular
That makes absolutly no sence. The whole reason it has that code name is because of the GPU's in it. Although I see what your trying to go at, but with all due reason this is the new trend.
What?? If Rv770 is about 50% faster than r670 how could you get 200% when r700 will be 2xRv770?? It seems it will be about 2 times faster not 3
Yes three times faster than R600/RV670. RV770 = 1.5 of RV670 correct? Going off of performance rumours of course. That would imply R700 = 1.5 + 1.5 = 3x RV670. Again all assuming not being a AFR solution.