Video Card Characteristics that matter most
Most novice video card buyers are overwhelmed with many new terms used in Video Card world such as core clock, GPU, pixel rate, shader and others. Weíll guide you what things to pay most attention to when searching for a video card.
GPU Core Clock
Have you heard of a CPU (the Central Processing Unit)? You know how some CPUs are faster than others? Well, for video cards they call it a Graphics Processing Unit (or GPU). The frequency of this GPU is referred to as the Core Clock and is measured in megahertz (MHz). The higher this frequency value is the faster the video card can process the data. However, GPU is not the only important part of a graphics card that you need to pay attention to.
Most important characteristics of video memory are size, speed (frequency) and memory bus. A wider memory bus leads to more data transferred per second between the video card GPU and video memory. Modern video cards have memory sizes ranging from 128 MB to 1024 MB. Low-End and Mid-End products usually have 256 MB of memory. High-End come with 512 MB and more. The more memory is installed on a video card, the more data it can store, so there is no need to use slower access to PC RAM. If you have to choose between higher memory size vs. wider memory bus, go for the wider bus. Frequency is another parameter that directly affects memory bandwidth. Memory bandwidth depends both on memory frequency and its bus width and is very important factor in gaming and video rendering. For example, a video card with memory frequency of 1000 MHz with a 256-bit bus will have better bandwidth than 1400 MHz memory with a 128-bit bus. For serious gamers we donít recommend any video cards with less than 128-bit bus.
All modern video cards have DDR or GDDR memory. Itís usually specified by its effective frequency, which is doubled from the actual physical frequency. For example, if DDR frequency is specified as 1500 MHz, memory physically operates at 750 MHz.
A video card draws every image on the computer monitor by filling the screen with dots. The fillrate is how fast the video card can render and Ďfillí the screen image with the dots. For serious 3D gamers this is very important since the screen image changes very fast and usually contain many details. In order to experience games at their peak levels, your card must have the ability to sustain an extremely high pixel fill rate.
There are two types of fillrate: pixel fillrate and texel rate. Pixel fillrate characterizes the speed of drawing pixels and depends on the operating frequency and an amount of ROP (raster operations pipeline) units. The latter is the texture fetch rate, which depends on the operating frequency and an amount of texture units.
For newer cards the problem bottleneck has shifted away from fillrates to pixel shader.
There are three forms of shaders: pixel shaders, vertex shaders and geometry shaders (later introduced with DirectX 10 and OpenGL 1.1). They are all software programs running on a Graphics Processing Unit (GPU).
Pixel shader is a software program written to manipulate pixel color, usually to apply an effect on an image such as shadows effect or realism effect. It implements a complicated graphics algorithm that calculates effects on a per-pixel basis.
Vertex shader is a program which deforms 3D elements. It usually adds special effects to 3D objects.
Geometry shader is a program which generates new graphics primitives, such as points, lines and triangles. Geometry shader programs run after vertex shaders but before pixel shaders. Typical uses of a geometry shader include point sprite generation, shadow volume extrusion, and single pass rendering to a cube map.
Next: Expansion Slot