Protein database search using Processing-in-Memory architecture

This paper describes the implementation of alignment search within a protein sequence database using the Processing-in-Memory concept. The PiM memory component used is a specialized DRAM developed by UPMEM. The study examines two main aspects: the acceleration achieved compared to a traditional multi-core server and the energy savings enabled by this architecture. These evaluations highlight the potential of PiM technology for enhancing the performance and efficiency of computationally intensive bioinformatics applications.