Several diseases, such as sickle cell anaemia, malaria, cancer, etc., are caused by pathological action that alters the mechanical properties of healthy red blood cells (RBC). Understanding the behaviour of healthy and diseased cells will help to develop diagnostic devices for the early detection of diseases. Lateral migration is a cell separation technique in which the cells can be separated from different lateral positions based on their size and deformability resulting from the migration of cells across streamlines under the action of hydrodynamic forces. In the present work, a numerical model based on smoothed particle hydrodynamics has been developed to study complex fluid-structure interaction of RBC and parallelized in GPU, leading to a seventeen-fold reduction in computational cost. RBC is assumed to be circular, and a spring network model is used to model the RBC membrane in the present work. The developed numerical model is used to investigate lateral migration and deformation of malaria infected RBC in Poiseuille flow and compared with that of healthy RBC.