The full-featured graphics-enabled mode of VMD is the most demanding, and requires an OpenGL-capable graphics accelerator with up-to-date drivers. Some graphics chipsets or GPUs come with drivers that are below-spec and will not be able to run VMD with full graphics capability. These will either automatically, or as the result of user-defined environment variables (e.g. VMDSIMPLEGRAPHICS), use a reduced functionality graphics mode within VMD. Since the choice of the GPU chipset or card has the biggest impact on the visualization capabilities and performance of VMD, this is the hardware component that is worth spending money on if one's intended use of VMD is primarily focused on visualization related tasks. VMD implements a variety of advanced rendering features that hinge upon the availability of GPU hardware and driver support. When available, these features enable VMD to interactively display very large or complex structures and support a variety of special stereoscopic 3-D displays [1,2,3,4,5,6,7]. As an added bonus, recent GPUs are now also capable of accelerating some of the computationally demanding tasks within VMD, discussed in more detail below.
Following the choice of graphics accelerator, the amount of available system memory tends to have the next most significant impact on the performance and capability of VMD. The more memory a machine has, the more frames can be loaded at once from large molecular dynamics trajectory files. For batch-mode analysis tasks that consist primarily of scripting, system memory is frequently the resource that limits feasability of many analysis tasks.