GPU cards have their own built-in memory and can contribute to the resolution of complex computing tasks. The additional built-in processing power in GPUs relieves task-intensive strain on the CPU, enabling your application to run faster. By default, the software automatically detects and uses the GPU if one is available on the computer and it is able to deliver a speed enhancement compared with using the CPU. The GPU options setting is found in the solver parameters dialog.
Parallel solution technology improves solution speed. It is implemented, as an option, in the solver parameters, and turned on by default, but you can be turn it off if you prefer. It allows major computational tasks to be computed simultaneously by a team of threads on a processor, or many processors, when running an analysis. GPU technology, and our parallel solver complement each other; if you use GPU technology in conjunction with a parallel solver, you can achieve even greater speed improvements.
These features have been available for Windows operating systems for several years. Additional enhancements to these features are listed below:
A new option, Automatic, has been added for this release, where the solver determines the most efficient number of threads to use, based on the CPU usage.