Save computational time – the PARALUTION Library provides state-of-the-art iterative sparse linear solvers, which can utilize modern multi-core CPU and accelerator technologies such as Graphics Cards (GPU) and other accelerator (MIC/Xeon Phi) architectures. Based on our software, the user can significantly decrease the computation time of the performed simulations. Special focus is on computational fluid dynamics (CFD), electromagnetics and stress analysis, where PARALUTION provides advanced solvers which can be embedded into well-known software packages and thus help to accelerate time-critical simulation parts and save time.
Save R&D time – the syntax and the structure of the library provide an easy learning curve. With the help of the provided examples, anyone can try out the library and program GPUs – no knowledge in CUDA, OpenCL or OpenMP required. Thanks to the portability, a software developer that uses the PARALUTION Library can develop code using cheap, standard hardware. Once she/he has finished, she/he can move the finalized program, which is based on the PARALUTION Library, to any hardware (that is supported by PARALUTION), and execute her/his program. No code modifications will be necessary. The software developer will obtain identical results to her/his cheap, standard hardware computer when he executes her/his program on a (GPU) system. Only computation performance will be noticeably different.