Using Ram over VRAM

[brencmcd]

Currently getting my feet wet with Zenith simulations and on my machine I have much more Ram available over my limited VRAM. I am loving Zenith so far but am constantly running into "Max VRAM limit reached". Curious about what options are currently in Zenith that can allow for better use of my Ram over VRAM during intense simulations.

[tyFlow]

Zenith is a GPU-only solver...it'll "use" RAM if CUDA spills over to RAM once VRAM is full, but that can lead to extreme simulation slowdowns (once RAM swapping happens inside CUDA, everything becomes orders of magnitude more inefficient). If you'd like to enable spill over like that you can disable the VRAM limit in the Birth Inferno operator.

Beyond that, the solution is to invest in a GPU with more VRAM