The guys at The Enquirer seem to have some dough about the next generation GPU from  nVidia , In an analyst webcast, Nvidia Investor Relations and Communications VP Michael Hara stated the top end card based on the G92 graphics processor will be ready for Christmas and that it will have computing power close to 1 TeraFLOP .
Time for some cools stats now , Intel’s latest quad core Core 2 extreme QX6700 runs at 2.6 GHz and has a peak floating point performance of 50 GFLOPS, while nVidia’s G80 ( powers the 8800GTX)  has a peak floating point performance of 330 GFLOPS, AMD’s R600 can do a max 450 GFLOPS  and the STI Cell B.E does close to 250 GFLOPS at max throughput . So if the G92 rumor holds it would mean that the G92 will outperform the cell by around 4 times and its predecessor G80 by about 3 times  in terms of computing power.
One problem with GPU’s  is the inability to support 64 bit floating point operations, which is necessary for almost all supercomputing applications . On the other hand GPU’s are far cheaper than supercomputer vector processors making their use in HPC( High performance computing) justified . nVidia had promised FP64 support on GPUs by late 2007 , So is G92 the promised chip , if so its going to have a big impact on the GPGPU movement and on HPC in general.
One little concern for me is the lack of titles for the PC which can take advantage of such horsepower , Crysis is a contender and so is SC-Conviction . TeraFLOP or not it’ll be interesting to see what kind of performance can the G92 can dish out.

nVidia has strapped on 2 Quadro FX5600 GPU’s ( OpenGl version of the mighty 8800GTX) onto thier already impressive line of Quadro plex VCS ( Visual computing system) . Looks like they are getting ready to take on the emerging GPCPU market as the new workstations support GPU computing with nVidia’s CUDA programming.
GPCPU has become a lucrative market for the future and a lot of big names are pouring money into it , recently Rapidmind , a startup grabbed $10 million of VC money for their GPCPU platform, keeping in pace nVidia is pushing its graphics cards hard as a platform for massively multi-threaded processing applications.
With the new GPU’s, the total frame buffer goes up to 3Gb (1.5 Gb per GPU) FSAA (full screen anti-aliasing) goes up to 64X , and as with the 8800GTX , the GPU’s have a unified shader architecture and fully support shader model 4.0 . Now comes my favorite part – performance stats 64x SLI FSAA,16 synchronized output channels,8 HD SDI channels,60 billion pixels/sec fill rate,1 billion triangles/sec geometry performance and is able to handle up to 148 megapixels display walls. and you can have many of these Quadro Plex boxes in your visualization cluster for scalability. Oh ! I forgot to mention they cost $18000 a piece.


With all the hype surrounding vista and with the ongoing console wars , The Nvidia 8800 GTX launch went almost unnoticed , I am not lucky enough to own one but i surely gobbled up the demo videos on you-tube , Crysis running on max settings , and the stunning UT3 are just some examples of what this graphics powerhouse can do .
Unfortunately at the Vista launch there were just two DX10 cards available , the Nvidia 8800GTX and the scaled down version of the same 8800GTS , somewhere around march ,I expect Nvidia will start shipping mainstream and lower end DX10 products based on the G86 and G84 , this most probably will coincide with the ATI R600 launch which promises to be the most advanced graphics processor ever to grace this earth , by April ATI will announce their lower end DX10 range of GPU’s and then we have the classic Red vs Green battle again for the rest of the year.
A direct impact of this will be on the pricing of the current generation of GPU’s , prices are expected to fall substantially and since it will take some time for DX10 to settle in properly , it looks like the right time to invest in a top of the line DX9 GPU .