Possible specs for Nvidia's G80 leak out
Some reasonably detailed specifications along with pricing and naming information for Nvidia's upcoming G80 graphics processor have leaked. The specs were posted by VR-Zone but have since been silently taken down. Luckily, the information was relayed by the folks at X-bit labs and reads as follows:
- Unified Shader Architecture;
- Support FP16 HDR+MSAA;
- Support GDDR4 memories;
- Close to 700M transistors (G71 - 278M / G70 - 302M);
- New AA mode: VCAA;
- Core clock scalable up to 1.5GHz;
- Shader peformance: 2x Pixel/12x Vertex over G71;
- 8 TCPs & 128 stream processors;
- Much more efficient than traditional architecture;
- 384-bit memory interface (256-bit+128-bit);
- 768MB memory size (512MB+256MB)
The information says the G80 will launch in November in the form of GeForce 8800 GTX and GeForce 8800 GT models priced at $649 and $449-499, respectively. The faster of the two will have a 384-bit memory interface, a hybrid water and air cooler, and "7 TCPs." The GeForce 8800 GT will reportedly have a 320-bit memory interface, a standard air cooler, and "6 TCPs."
Neither X-bit labs nor VR-Zone clarify the meaning of the TCP acronym, but it could refer to thread control processors. According to this Nvidia patent, thread control units are used to configure programmable computation units to perform vertex or pixel operations. In other words, they set the function of the graphics chip's unified shader processors.