|
NVIDIA GeForce 7800GTX
Intro 2
Architectural Improvements The GeForce 7800 GTX contains 8 vertex units and 24 pixel pipelines. Operating with a core clock of 430MHz and a memory clock of 600MHz (1200 mega-transfers), and using 256-bit GDDR3, memory bandwidth is a whopping 38.4GB/sec. The 24 pixel pipes provide an astounding bilinear filtered texel fillrate of 10.32Gtexels/sec! And the GeForce 7800 GTX includes 128-bit floating point (32 bits/component) through the entire graphics pipeline. Compared to the GeForce 6 series, the internal pipelines were redesigned for lower latency. The Instructions per clk is also improved. Vertex units are reengineered to speed up geometry processing and new texture engine for texture processing. New transparent super sampling and multisampling antialiasing modes provide more realistic rendering of scenes otherwise impossible to filter well with standard supersampling and multisampling techniques NVIDIA High Dynamic-Range (HDR) rendering uses 16-bit floating point components (64-bit FP total) to produce incredible dynamic range of scene lighting. Dark objects appear really dark, and bright scenes or lights can be very bright, with visible details still present at both extremes as well as the gradients in between. Developers can easily convert their applications to use HDR rendering and add many special effects not possible with standard 32-bit (8-bit per component) RGBA representations. (See the Far Cry images in the benchmark section below for HDR examples, in addition to the PowerPoint presentation in your reviewer’s kit showing Splinter Cell Chaos Theory screen shots.) Next >>> |
(C) Copyright 1998-2009 OCWorkbench.com
|