Video random access memory

From Wikipedia, the free encyclopedia
GDDR5X SDRAM on an NVIDIA GeForce GTX 1080 Ti graphics card

Video random access memory (VRAM) is dedicated computer memory used to store the pixels and other graphics data as a framebuffer to be rendered on a computer monitor.[1] This is often different technology than other computer memory, to facilitate being read rapidly to draw the image. In some systems this memory cannot be read/written using the same methods as normal memory; it is not memory mapped.


Independent system RAM and video RAM
Unified memory
A GPU die surrounded by VRAM chips

While a computer has system RAM, most contemporary graphics cards have access to a dedicated set of memory known as VRAM. In contrast, a GPU which shares system memory has a Unified Memory Architecture, or shared graphics memory.

System RAM and VRAM has been segregated due to the bandwidth requirements of GPUs,[2][3] and to achieve lower latency since VRAM is physically closer to the GPU die.[4]

Modern VRAM is found in a BGA package[5] soldered onto the graphics card.[6] Like the GPU itself, the VRAM is cooled by the GPU heatsink.[7]


See also[edit]


  1. ^ Foley, James D.; van Dam, Andries; Feiner, Steven K.; Hughes, John F. (1997). Computer Graphics: Principles and Practice. Addison-Wesley. p. 859. ISBN 0-201-84840-6.
  2. ^ "What is VRAM: The Memory Power Behind Real-time Ray-Tracing".
  3. ^ "Relationship Between RAM and VRAM Bandwidth and Their Latency".
  4. ^ "RAM vs. VRAM: What's the Difference?".
  5. ^ "Encapsulated in CPUs, GPUs, RAM and Flash: Types and Uses".
  6. ^ "Graphics Card Components & Connectors Explained".
  7. ^ "Different Types of Graphics Card Cooling Solutions for GPU, VRAM & VRM".
  8. ^ "VRAM vs RAM | Differences & Applications". TechDim.
  9. ^ "GPU Framebuffer Memory: Understanding Tiling".