Handwritten signature has a key role in personnel authentication today. Institutions such as banks are using the handwritten signatures to control false transactions. Due to an increase in access by people to institutions such as banks, insurance companies etc., the volume of signatures are increasing day by day and need to be verified in real time. Hence, there is always a demand for a reliable and fast signature verification system. Researchers have contributed several techniques and ideas for signature verification. However, it has been investigated that the models are inefficient and their real-time implementation is a difficult task. In the present manuscript, a GPU based offline signature verification model has been proposed. The proposed signature verification model uses Hidden Markov Model (HMM) for both training and verification of the signatures. It has been found that the proposed GPU based model has more than 5 times faster response. The quantitative evaluations have been done using the average false rejection rate (FRR) and false acceptance rate (FAR). The lower values have been obtained which indicates that the proposed.