Genomic co-processor for long read assembly

Genomics data is transforming medicine and our understanding of life in fundamental ways; however, it is far outpacing Moore's Law. Third-generation sequencing technologies produce 100X longer reads than second generation technologies and reveal a much broader mutation spectrum of disease and evolution. However, these technologies incur prohibitively high computational costs. In order to enable the vast potential of exponentially growing genomics data, domain specific acceleration provides one of the few remaining approaches to continue to scale compute performance and efficiency, since general-purpose architectures are struggling to handle the huge amount of data needed for genome alignment. The aim of this project is to implement a genomic-coprocessor targeting HPC FPGAs starting from the Darwin FPGA co-processor. In this scenario, the final objective is the simulation and implementation of the algorithms described by Darwin using Alveo boards, exploiting High Bandwidth Memory (HBM) to increase its performance