A high performance computing for AOM stock trading order matching using GPU

Ketsarin Rungraung,P. Uthayopas
DOI: https://doi.org/10.1109/ICSEC.2013.6694750
2013-09-01
Abstract:The task of trading orders matching in financial markets is a very challenging task since due to the speed of arriving request. In this paper, the GPUs technology and CUDA programming is explored as a potential technology to accelerate this task. The trading method in Automatic Order Matching (AOM) of Stock Exchange of Thailand (SET) is used as a case study. The code is developed and parallelized using GPU technology to speeding up the. The experimental results indicated that GPU can increase the performance for this process. In addition, the speedup of the work increases with the data size. Thus, the use of GPU can be substantially important when there is a need to handle massive data at a very high speed.
What problem does this paper attempt to address?