I can not believe I missed the announcement of nVidia's G80 specifications -- that is the new Geforce 8800 series. Ok, so DailyTech have reported 3 articles on the GeForce 8800GTX and 8800GTS. So I will attempt to post them all in one news article and add some comment. The only bad news is we do not have any leaked benchmarks for the G80; hopefully, they will come soon enough.

Right first the important specifications:

GeForce 8800GTX

DirectX 10 Compatible
128 "Streaming processors" @ 1350Mhz
768 MB of Video RAM
575 Mhz core, 900 Mhz Memory
384-bit Memory Interface
Dual-slot cooling solution.

GeForce 8800GTS

DirectX 10 Compatible
96 "Streaming processors" @ 1200Mhz
640 MB of Video RAM
500 Mhz core, 900 Mhz Memory
320-bit Memory Interface
Dual-slot cooling solution.

Both the GeForce 8800GTX and 8800GTS products will be HDCP compliant with support for dual-link DVI, VIVO and HDTV outputs. nVidia expect the GeForce 8800GTX and 8800GTS products to launch the second week of November 2006, and it should be a hard launch -- much like the 7800GTX's launch.

Then in a second article DailyTech revealed more juicy news about the G80 cards.

The G80 will feature 128-bit high dynamic-range (HDR) and anti-aliasing with 16X sampling. Or in full:

* Higher dynamic range and color precision than existing 8- and 10-bit image file formats.

* Support for 16-bit floating-point, 32-bit floating-point, and 32-bit integer pixels. The 16-bit floating-point format, called "half", is compatible with the half data type in NVIDIA's Cg graphics language and is supported natively on their new GeForce FX and Quadro FX 3D graphics solutions.

* Multiple lossless image compression algorithms. Some of the included codecs can achieve 2:1 lossless compression ratios on images with film grain.

* Extensibility. New compression codecs and image types can easily be added by extending the C++ classes included in the OpenEXR software distribution. New image attributes (strings, vectors, integers, etc.) can be added to OpenEXR image headers without affecting backward compatibility with existing OpenEXR applications.

NVIDIA already has 16X AA available for SLI applications. The GeForce 8800 will be the first card to feature 16X AA on a single GPU. Previous generations of GeForce cards have only been able to support 8X antialiasing in single-card configurations.

DailyTech also revealed information about the power requirements. NVIDIA claim the card requires at least a 450W power supply for a single GeForce 8800GTX, and 400W for the 8800GTS. In addition, vendors in Taiwan have confirmed with DailyTech that GeForce 8800 cards in SLI mode will likely carry a power supply "recommendation" of 800W. These are some incredible requirements, and it seems nVidia have given up the low power requirements of the 7800 series for raw power. This probably is due to the 90nm process, so expect the refresh product to be 80nm, or even 65nm, and use a lot less power.

Finally documents have been leaked which show nVidia plan to use the G80 for physics processing. Around the time of the 8800 series launch NVIDIA will release a new engine dubbed Quantum physics engine. This Quantum Effects Technology is akin to NVIDIA's PureVideo Technology; it is a dedicated layer on the GPU dedicated to physics calculations. So it seems nVidia are attempting to nail Aegia's coffin shut.

All we need now is some concrete R620 news...