What will replace NV50?

bbot

Regular
Does the cancellation of the NV50 have something to do with the Nvidia's announcement that it will collaborate with Sony to design a custom gpu for the PS3? Does anyone know what Nvidia is planning to replace the NV50 with?
 
Well...I don`t know what they plan but i think that their next GPU will be faster than r520...Do yoy remember situation with NV40???I think that NVs next GPU will be as revolutionary and fast as NV40 when it was released...
 
Is the next generation of Nvidia cards likely to be delayed due to PS3 (due to diverting all of their resources onto that)? Perhaps they will skip a generation, and release their PS3 generation technology on PC graphics cards some time closer to the release of the PS3...
 
I have been long pretty much quiet, but now I think it's time say something.

uh...
I drop by on coffee break to read B3d and 2 first threads I found are fan boys discussing about shall their favourite company only beat up competition or shall they crush everyone and own the market... :rolleyes:

luckily the worst one was locked and I am sure that this one will be too. (I know moderators are right now on edge of question: "lock or not lock, that's the question."

and back on the subject... cancellation of what? did nvidia announce they cancelled it? did they even stated that there was even one? How you can cancel that does not exist?

this is redicilous discussion. It's just same as I claiming that AMD cancelled K16.

Also, did making chip to XBox forced to nvidia focus less on PC market? I don't think so. Why this new deal with PS3 would either. Also, I haven't read anywhere that nVidia would be directly doing GFX core to PS3. Only thing announced is that they are co-operating with Sony on this particular matter. (just licensing something small to sony, like DVI-interface would be co-operating and Of course because Sony is a giant and known of NIH thinking, nvidia certainly tries to catch some marketing value out of the deal, even it would be considering something like few resistors on the PCB.)


so, hold on your horses. too much speculation is too much speculation, even in GFX business.
 
But speculation is fun! Although most will be inaccurate, the odd nugget does turn out to be correct.
 
Nappe1 said:
I have been long pretty much quiet, but now I think it's time say something.

uh...
I drop by on coffee break to read B3d and 2 first threads I found are fan boys discussing about shall their favourite company only beat up competition or shall they crush everyone and own the market... :rolleyes:

luckily the worst one was locked and I am sure that this one will be too. (I know moderators are right now on edge of question: "lock or not lock, that's the question."

and back on the subject... cancellation of what? did nvidia announce they cancelled it? did they even stated that there was even one? How you can cancel that does not exist?

this is redicilous discussion. It's just same as I claiming that AMD cancelled K16.

Also, did making chip to XBox forced to nvidia focus less on PC market? I don't think so. Why this new deal with PS3 would either. Also, I haven't read anywhere that nVidia would be directly doing GFX core to PS3. Only thing announced is that they are co-operating with Sony on this particular matter. (just licensing something small to sony, like DVI-interface would be co-operating and Of course because Sony is a giant and known of NIH thinking, nvidia certainly tries to catch some marketing value out of the deal, even it would be considering something like few resistors on the PCB.)


so, hold on your horses. too much speculation is too much speculation, even in GFX business.

I agree with you but disagree at the same time. Yes Nivida did take their eye off the ball with Xbox. I believe this is how the got caught releasing NV30 while ATI released its R300. At least with Xbox the arcitecture was similar to the PC, so with PS3 I wonder how much of the fab and tech they will be able to convert over to the PC.
As for speculation, what would be the purpose of these forums without specualtion??
 
Nappe1 said:
so, hold on your horses. too much speculation is too much speculation, even in GFX business.

Huh? Neither the thread title nor the first post was fanboyish in any way and speculation has always been rife on this board. Maybe you should have had that coffee before you posted :LOL:
 
What will replace NV50?
Ok, everyone together now:

"The nV60!"

rofl.gif
rofl.gif
rofl.gif


I can't believe no one posted that yet! :LOL:
 
ninelven said:
I'd guess they are replacing nv50 with nv50.

Exactly what I thought. Why not, nobody else knows what was originally planned. Any project can grow/change at any time, but you don't need to change names.
 
ninelven said:
I'd guess they are replacing nv50 with nv50.

That would be locical from an external perspective, but it all depends on how the codenames are used internally. When you start down the path of developing something you'll have all kinds of different code threads (hardware and software) that relate to that codename; if you change the roadmap to a different product / development composition it can become very difficult to separate the the "old" development code from the "new" development code if they are using the same codenames. Its this reason why R420 was called R420, not R400 as they already had considerable work attributed to R400.

Of course the codenames are getting nearly as well known as the product names these days so it may be possible that even the chip codenames may not relate to the internal development names anymore.
 
maybe the NV50 is scrapped in favor of NV55 or what Nvidia would have done with NV55.

or maybe Nvidia is bringing its R&D from NV60 down to the NV5X generation?
 
Maybe Nvidia will go for a single board SLI solution, like this:

http://www.tomshardware.com/hardnews/20041216_115811.html

These are only 6600GTs, but imagine some beefed up 6800s. I know Nvidia said they wanted to push SLI, maybe this is their solution to the longer development cycles. Instead a new family every 18-24 months with a refresh every 9-12mo, why not do a single board SLI 12mo after release so they can focus all their energy on the new family?

I know it is a wild idea, but it seems that in between DirectX updates, as long Nvidia was able to do all the new features well, this could be a solution to the ever increasing development costs. I know it is probably too expensive, but who knows... just a wild thought. One I am pretty sure is wrong :)
 
Nappe,

While in general I don't disagree with you, this one call could for some more discussion:

Also, did making chip to XBox forced to nvidia focus less on PC market? I don't think so. Why this new deal with PS3 would either. Also, I haven't read anywhere that nVidia would be directly doing GFX core to PS3. Only thing announced is that they are co-operating with Sony on this particular matter. (just licensing something small to sony, like DVI-interface would be co-operating and Of course because Sony is a giant and known of NIH thinking, nvidia certainly tries to catch some marketing value out of the deal, even it would be considering something like few resistors on the PCB.)

It is true that at this stage little is known about what exactly NV will deliver for the PS3. Apart from that though all companies that gained console deals, had to divert resources and make internal changes in their roadmaps.

XBox1/NV2a was part of the reason why NV25 and NV30 had been postponed further into the future (where it's not clear either that the "original" NV30 had that much in common with the final released result) and ATI had to "scratch" in relative terms their R400 and had a few shifts and codename changes in it's own roadmap.

I don't see why things should be different at NVIDIA. The real question would be (see also Dave's last post) what each codename really stands for, what changes and where have occured to the roadmap and when that shift actually happened.

If you'd ask the INQ for example what they exactly think the "NV50" codename stands for, they'll most likely just lift their shoulders in ignorance.

Bottomline for me is that neither IHV will release a "WGF2.0" compliant chip prior to H2 2006 (from the current expectations of a Longhorn release) and that anything between then and today CANNOT be anything than another SM3.0 based chip. Now happy speculators can attach codenames and possible specs as their heart desires, I have severe doubts that the essence of this very paragraph will change a bit.
 
Could it be that say NV50 was an AGP part and NV55 would be the PCI-E version of the same part similar to NV40 and NV45? Cancelling NV50 might then not mean much as they could use NV55 with a bridge chip for AGP.
 
Back
Top