In the old days, before 3D accelerators, VRAM was just where the picture you saw on the screen was stored. This is called a framebuffer.
It was a separate RAM chip, with a neat additional feature: it was “dual-ported,” meaning you could read/write to it from two “sides”.
One side was the computer itself: you can read and write to it just like any other RAM in your computer. Just at another address than the rest of your RAM
The other side was the monitor output: it would just read the VRAM, and put it out as an analog signal, 60 times per second (or whatever the refresh rate was set to).
And nowadays with 3D accelerators, it got of course a bit more complicated. VRAM now also contains e.g. textures for 3D objects. Modern graphics card have their own processors (actually a whole lot of them… up to 18’000 of them). They can do stuff like applying textures in VRAM directly to 3D objects in the framebuffer. And the framebuffer is now just a small part of the whole VRAM.
Latest Answers