general purpose graphics processing unit (GPGPU)

The rapid development of graphics processing units( GPU) and their software has led to these high- performance processors being increasingly used for computationally intensive tasks in high-performance computers, high-performance clusters( HPC) and physics engines. Therefore, in this context, one speaks of General Purpose Graphics Processing Unit (GPGPU).

The universal use of graphics processors in compute-intensive graphics calculations is supported by Compute Unified Device Architecture (CUDA), an Nvidia development, with which suitable applications that run in C++++ on central processing units( CPU) can be ported to graphics processors. The graphics processor thus becomes the co-processor of the centralprocessingunit and relieves it of computationally intensive tasks. For corresponding applications, graphics processors have significant advantages over central processing units in terms of computing power and investment.

GPGPUs are particularly suitable for tasks that consist of many uniform individual calculations. They are always advantageous when CPUs are to be relieved of graphics calculations. For example, in geological, seismological and geographical calculations, in medical applications and in simulations and crash tests.