high performance computing on graphics processing units: hgpu.org

This paper deals with using tools commonly available to programmers to implement the finite difference time domain (FDTD) calculations using video cards. In the past few years developments in the field of graphic processing units (CPU’s) for video cards have vastly outpaced their general central processing unit (CPU) counterparts. As specifically applied to vector mathematic operations, the newest generation GPU’s can generally outperform current CPU architecture by a wide margin. With the addition of large onboard memory units with significantly higher memory bandwidth than found in the main system, graphic cards can be utilized as a highly efficient vector mathematic co-processor. Implementing functions in high-level languages that utilize the vector processing power of the video cards, an appreciable increase in the effective speed for vector and matrix computations occurs in common FDTD implementations can be achieved. By formulating proper procedures to realize general vector computations on GPU’s it will be possible to maximize the processing power available to an extent greater than possible without the addition of the video card.