Mr. Soratouch's research presentation at PDCAT 2023

Mr. Soratouch Pornmaneerattanatri, a second-year doctoral student at SDLab presented the following paper at the 24th International Conference on Parallel and Distributed Computing, Applications and Technologies (PDCAT2023)

Soratouch Pornmaneerattanatri, Keichi Takahashi, Yutaro Kashiwa, Kohei Ichikawa, Hajimu Iida, “Parallelizable Loop Detection using Pre-trained Transformer Models for Code Understanding”, 24th International Conference on Parallel and Distributed Computing, Applications and Technologies (PDCAT2023), August. 2023.

In this paper, we proposed an alternative methodology for automatic parallelization tools to classify parallelizable For-loops using a deep learning-based natural language processing model. In this approach, we improved the accuracy of the automatic parallelization tool in identifying parallelizable For-loops compared to the traditional static analysis approach by a fine-tuned CodeT5 model.