Use oneAPI to Port a Framework and Code to FPGAs
Whole-genome sequencing is a key in genomics. Since sequencing takes a long time to run, it is accelerated by applying calculations, parallelization, and offloading with a GPU.
The most widely used genome analysis toolkit (GATK) is written in Java*, so it is impossible to directly use oneAPI for porting it to an FPGA. But, the toolkit can be ported during the aligning and mapping step, which happens before porting. This step is written in C/C++ and uses SYCL* for the Burrows-Wheeler Aligner maximal exact matches (BWA-MEM) and DRAGEN Map and Align (DRAGMAP) programs.
Speaker
Jinwoo Jung has 20 years of experience in HPC that includes code optimization and parallelization, tech support, and training. He is an Intel oneAPI instructor, and is researching how to apply HPC and AI with Intel oneAPI.
Product and Performance Information
Performance varies by use, configuration and other factors. Learn more at www.Intel.com/PerformanceIndex.