Compare any two graphics cards:
GeForce GTX 460 1GB vs Radeon HD 4890 2GB
IntroThe GeForce GTX 460 1GB uses a 40 nm design. nVidia has set the core speed at 675 MHz. The GDDR5 RAM is set to run at a speed of 900 MHz on this specific card. It features 336 SPUs along with 56 TAUs and 32 ROPs.
Compare those specifications to the Radeon HD 4890 2GB, which features a core clock frequency of 1000 MHz and a GDDR5 memory frequency of 975 MHz. It also features a 256-bit bus, and makes use of a 55 nm design. It is made up of 800(160x5) SPUs, 40 Texture Address Units, and 16 ROPs.
Power Usage and Theoretical Benchmarks
Power Consumption (Max TDP)
In theory, the Radeon HD 4890 2GB will be 8% quicker than the GeForce GTX 460 1GB in general, because of its higher data rate. (explain)
Texel RateThe Radeon HD 4890 2GB will be a little bit (more or less 6%) better at AF than the GeForce GTX 460 1GB. (explain)
Pixel RateIf running with lots of anti-aliasing is important to you, then the GeForce GTX 460 1GB is a better choice, and very much so. (explain)
Please note that the above 'benchmarks' are all just theoretical - the results were calculated based on the card's specifications, and real-world performance may (and probably will) vary at least a bit.
Memory Bandwidth: Memory bandwidth is the maximum amount of information (in units of megabytes per second) that can be moved across the external memory interface within a second. It is calculated by multiplying the bus width by the speed of its memory. If it uses DDR type memory, the result should be multiplied by 2 again. If DDR5, multiply by 4 instead. The higher the memory bandwidth, the better the card will be in general. It especially helps with AA, High Dynamic Range and high resolutions.
Texel Rate: Texel rate is the maximum number of texture map elements (texels) that are processed per second. This figure is worked out by multiplying the total texture units by the core speed of the chip. The better this number, the better the graphics card will be at handling texture filtering (anisotropic filtering - AF). It is measured in millions of texels in one second.
Pixel Rate: Pixel rate is the maximum amount of pixels the graphics card could possibly record to the local memory per second - measured in millions of pixels per second. Pixel rate is calculated by multiplying the amount of Render Output Units by the clock speed of the card. ROPs (Raster Operations Pipelines - sometimes also referred to as Render Output Units) are responsible for outputting the pixels (image) to the screen. The actual pixel rate also depends on lots of other factors, especially the memory bandwidth - the lower the memory bandwidth is, the lower the ability to reach the maximum fill rate.