Enabling Faster NGS Analysis on Optane-Based Heterogeneous Memory

Authors: Jiaolin Luo, Luanzheng Guo, Jie Ren, Kai Wu, and Dong Li (University of California, Merced)

Abstract: Next-Generation Sequencing (NGS) analysis technologies are a pioneering approach for genome sequencing. The computation of NGS analysis exhibits a unique pattern, in which the execution requests a high density of small I/Os in the process de novo genome assembly. The small I/Os can have a huge impact on performance and delineate sequencing performance. To solve the problem caused by small I/Os, we leverage the byte-addressable feature of emerging persistent memory, which has largely transformed the computation architectures of HPC. We first conduct experiments to study the impact of persistent memory on NGS analysis. Furthermore, we propose an optimization mechanism, which converts POSIX read/write calls to pure memory LOAD/STORE instructions at runtime, to significantly advance I/O efficiency. Our evaluation demonstrates the effectiveness of the devised optimization mechanism, in which we achieve a performance improvement of 31%.

Best Poster Finalist (BP): no

Poster: PDF
Poster summary: PDF


Back to Poster Archive Listing