Graphics core improvements

For their second-generation shader model 3.0 part, NVIDIA builds largely on the foundation first established with GeForce 6800 Ultra, improving key parts of the graphics core’s architecture to better handle today’s games and applications. For instance, NVIDIA is quick to boast that during the GeForce 7800 GTX’s development, they analyzed 1,300 of the most commonly used shaders in order to determine which features to focus on for GeForce 7800 GTX. As a result of this analysis, NVIDIA discovered that multiply-add (MADD) operations are one of the most commonly used math functions today. These operations are commonly used in lighting, normal map calculations, and many other operations.

Based on this, NVIDIA doubled the number of multiply-add instructions in each pixel pipe for GeForce 7800 GTX, increasing throughput for the pixel shader. On the vertex side, NVIDIA has tweaked the vertex shader units so that they can now process MADD operations in a single clock cycle. The vertex units were also enhanced to speed up geometry processing. For instance, triangle setup is over 30% faster. This shortens the time required to setup complex geometry processing, improving the throughput of the 3D engine. According to NVIDIA, the enhancements they’ve integrated into GeForce 7800 GTX increase pipeline efficiency by 50% on a clock-for-clock basis.

Note how much longer the 7800 GTX is than the other cards

Single-slot cooling

In addition to these tweaks, NVIDIA has also increased the number of pixel and vertex shaders. For GeForce 7800 GTX, NVIDIA has incorporated 24 pixel pipelines, eight more than the 16 pixel pipes found in GeForce 6800 Ultra. Like the GeForce 6800 Ultra, the 7800 GTX’s pipelines are arranged into quads, with each quad consisting of 4 pixel pipelines. The GeForce 7800 GTX consists of six quads (versus four in GeForce 6800 Ultra).

For vertex processing, NVIDIA increased the vertex unit count by two, with the GeForce 7800 GTX consisting of eight vertex shaders versus six in GeForce 6800 Ultra.

NVIDIA incorporated enhancements into more than just the pixel and vertex shaders. In order to improve texture performance, NVIDIA has improved texture fetching, allowing the graphics core to grab and access textures faster. Other improvements NVIDIA has incorporated include a 30%+ reduction in cycle time latencies in the fixed-function setup stages, and similarly cut cycles in gamma adjusted rotated grid AA. The 7800 GTX’s memory controller is more efficient too thanks to improved compression.

7800 GTX is much longer than X850 XT PE

In case all those enhancements aren’t enough, NVIDIA has also slightly increased the 7800 GTX’s clock speeds. NVIDIA’s reference specifications for the 7800 GTX call for a 430MHz graphics core (although we’ll delve into this more on the next page), and 600MHz memory. This is an improvement of 30MHz on the graphics core over the original 6800 Ultra’s specs, and 50MHz on the memory.