RSX: A G70 shrunk down to 90nm?
- Thread starter bbot
- Start date
By the time Nvidia and Sony told people about the deal they already had 18 months working on it,in fact this is say by Nvidia own people this put it with more time on the making that the Xenos.
http://www.nvidia.com/object/IO_17342.html
By december 7 2004 they say that they have 2 years working with Sony already,looking at how much time on the making the Cell has i'm not surprise,but yes they did have alot of time to work with the RSX,and they still do.
http://www.nvidia.com/object/IO_17342.html
By december 7 2004 they say that they have 2 years working with Sony already,looking at how much time on the making the Cell has i'm not surprise,but yes they did have alot of time to work with the RSX,and they still do.
gurgi said:
That specific quote? Not really. "Designing in parallel" doesn't mean they're the same - IBM designed Cell and X360 and Revolution's CPU in parallel, ATi developed R520 and R500 in parallel etc.
Of course, I do think they share the same parent technology at the very least, obviously. Like any generation of graphics chips, you have a series of derivatives, and RSX looks like a console derivative in this series.
I don't know how this thread can stir so much debate - haven't we been discussing this very topic for like a month now?
Anyway it's obvious RSX isn't going to be just a shrunk G70; that just wouldn't make any sense. PureVideo and pther crap has to go - and there must be some provisions in there for Cell interaction. That being said, how far it deviates from G70 - well we just don't know, maybe not much.
Those 3DMark scores were heavily effected by the CPU speeds it should be mentioned; somethign that's not typically the case in 3DMark.
Anyway it's obvious RSX isn't going to be just a shrunk G70; that just wouldn't make any sense. PureVideo and pther crap has to go - and there must be some provisions in there for Cell interaction. That being said, how far it deviates from G70 - well we just don't know, maybe not much.
Those 3DMark scores were heavily effected by the CPU speeds it should be mentioned; somethign that's not typically the case in 3DMark.
That doesn't mean they just wiggle thier nose and boom its shurnk down to 90nm .3roxor said:cause it has to be shurnk down to 90nm
But hasn't Sony already got some 90nm fab :?
I doubt its just a shrunk down g70. I think its most likely based on the refresh of the g70 which i think would come out early 2006 and it will have some improvements like fp32 hdr .
I highly doubt its anything more than a modified g75 (if you will ) It may even have a few less rops .
overclocked
Veteran
The way i see it is that because many that have been lurking on B3D for years are very interested in 3D gfx technology as a "whole".
To make a distant relation we can look at the current GS/NV2A and Flipper and see the discusion we have had about those chip and in many cases both the flipper and GS has been discused lively, and although the GS is the weakest/oldest its still in my mind probably the most interesting part because of its highly unortodox design, closely followed by flipper and then NV2A.
The reason IMHO is the same with RSX "vs" Xenos.
With that i mean saying that the RSX is a modified G70 and to what extent is impossible to say versus Xenos that is a glimpse in the future of the PC side and "more" custombuild for its purpose and radical design(thinking the Edram). These are the factors that i think triggers many here at B3D atleast that have a passion for 3d technology.
I think this is the same way the Cell cpu have many of us interested just because its a forward looking and custombuilt architecture.
But to make claims that one of the GPU´s is more powerful than the other is at this stage not valid from my POV.
To make a distant relation we can look at the current GS/NV2A and Flipper and see the discusion we have had about those chip and in many cases both the flipper and GS has been discused lively, and although the GS is the weakest/oldest its still in my mind probably the most interesting part because of its highly unortodox design, closely followed by flipper and then NV2A.
The reason IMHO is the same with RSX "vs" Xenos.
With that i mean saying that the RSX is a modified G70 and to what extent is impossible to say versus Xenos that is a glimpse in the future of the PC side and "more" custombuild for its purpose and radical design(thinking the Edram). These are the factors that i think triggers many here at B3D atleast that have a passion for 3d technology.
I think this is the same way the Cell cpu have many of us interested just because its a forward looking and custombuilt architecture.
But to make claims that one of the GPU´s is more powerful than the other is at this stage not valid from my POV.
Similar threads
