A graphics card is mainly composed of a processor, GPU (Graphic Processing Unit), and memory.
The graphics card is in itself a small PC dedicated to graphics applications, independent of the rest of the PC, the only link being the information passing through the port graphic and food.
Choosing a card is based on its needs and its budget:
- For desktop applications, a map called "entry level" enough;
- For multimedia applications, a map can read the desired video format is imperative;
- For 3D video games (and generally all 3D applications), one more powerful is essential.
Structure of a graphics card
The essential component that will determine the performance of the graphics card is the GPU:
A modern GPU is based on different units calculation:
- Processors Flow
- Texturing units,
- Raster units (ROPS).
The most important are the processors flows over the number of processors flow is important, the GPU will be more powerful (for architecture).
The model number on the card is linked to the GPU.
To enable the GPU to store its calculations, there are memory:
- It varies in quantity:
- 256 MB enough for the tickets range of previous generations;
- 512 MB is sufficient for most current cards graphics;
- 1GB is needed for cards with a very powerful GPU, the GTX 280;
- Broadly speaking, the more you play in high resolution and filtering enabled, you need more memory.
- It varies in type:
- DDR2, DDR3, DDR4 or DDR5: The higher the figure, the greater the memory is fast. His type does not relate to that used for RAM PC: you can have a PC in DDR2 and DDR3 graphics card.
- Its frequency:
- Plus it is high, the card is more powerful.
Finally, to link GPU and memory is the memory bus.
The width of this bus is important: it may be 64, 128, 256, 384 and 512 bits: over the bus is large, most trade between the GPU and memory will be rapid. A map to 256-bit bus is efficient at a map of 128-bit bus for the same GPU.
Bookmarks