Workstation Setup

Operating System

TurboWAVE can be readily installed on Linux, MacOS, or Windows. The workflow tends to be UNIX-oriented, which is natural for Linux and MacOS, and thanks to PowerShell, fairly seamless even on Windows.

Visualization tools are fairly universal, taking the form of command line Python scripts, or interactive Python notebooks. Outputs are written in Python-friendly formats so you can readily create your own post-processing tools.

Desktop OS’s that are explicitly supported and tested at the time of this writing include MacOS 10.15, RHEL 8.x and clones, Ubuntu 20.04, and Windows 10. Supercomputer platforms that are known to work well include Cray and SGI.

Hardware

At the time of this writing, the choice of CPU and memory capacity are the two major factors affecting turboWAVE performance, while choice of GPU is not very important. TurboWAVE is designed to effectively utilize multi-core processors. The number of cores is the most important performance hardware characteristic for turboWAVE. The more cores the better. The memory requirements depend most critically on the dimensions of the problem. A rule of thumb is to allow 1000 bytes of RAM for each grid cell. For a serious 2D problem this leads to about 1 GB. For a small 3D problem, several tens of GB are required.