VirtualRender

VirtualRender is yet another LightWave renderer. The difference from other solutions is that it does not use physical memory for framebuffer allowing to render the biggest imaginable scenes without using virtual memory, which slow down rendering a lot, or which immediately result in complete failure at initialization stage. That's why it's called "virtual" - rendered regions are saved immediately to hard disk without unnecessary taking memory during rendering rest of image.

VirtualRender is the first LightWave bucket renderer that splits framebuffer to small, user definable size regions. Each render node in your renderfarm now can work on the same frame of image telling other nodes which regions are in progress and which are rendered, and at the end there is done composition of all rendered regions into one with unlimited resolution, limitted only by available hard disk space. We were successfully rendering scenes with dual ScreamerNet setup on single machine which was taking 110 MB each node, 220 MB both, at resolution 16384x16384. LightWave renderer even failed to initialize. At 640x480 this scene was eating 800 MB of memory.