SWIFOLD: Smith-Waterman implementation on FPGA with OpenCL for long DNA sequences

Abstract

BackgroundThe Smith-Waterman (SW) algorithm is the best choice for searching similar regions between two DNA or protein sequences. However, it may become impracticable in some contexts due to its high computational demands. Consequently, the computer science community has focused on the use of modern parallel architectures such as Graphics Processing Units (GPUs), Xeon Phi accelerators and Field Programmable Gate Arrays (FGPAs) to speed up large-scale workloads.ResultsThis paper presents and evaluates SWIFOLD: a Smith-Waterman parallel Implementation on FPGA with OpenCL for Long DNA sequences. First, we evaluate its performance and resource usage for different kernel configurations. Next, we carry out a performance comparison between our tool and other state-of-the-art implementations considering three different datasets. SWIFOLD offers the best average performance for small and medium test sets, achieving a performance that is independent of input size and sequence similarity. In addition, SWIFOLD provides competitive performance rates in comparison with GPU-based implementations on the latest GPU generation for the large dataset.ConclusionsThe results suggest that SWIFOLD can be a serious contender for accelerating the SW alignment of DNA sequences of unrestricted size in an affordable way reaching on average 125 GCUPS and almost a peak of 270 GCUPS.

FPGA Acceleration of Sequence Alignment: A Survey

“…In [51] as an extension to [49], Intel Arria 10 is used as the hardware platform due to its incorporation in both new Xeon processors and the Intel-Go platform. They also provide a guide to choose the best platform for running SW for long reads.…”

“…In the previous works, the calculated alignment matrix had to be sent to the host to traceback the scores to find the aligned substrings, however, in this work all these steps are done in FPGA to further accelerate the algorithm. With the advent of FPGA architectures and providing more resources and introducing high-level synthesis tools, the work [51] uses OpenCL to reduce the complexity of implementing alignment algorithm on hardware. The work [65] supports queries without any limitation on the length…”

MASA-StarPU: Parallel Sequence Comparison with Multiple Scheduling Policies and Pruning

“…Related work in the area of Megabase DNA sequence comparison show that impressive performance can be attained with accelerators [3] [8] [9]. The performance results for multicores (CPUs) are less impressive but this is the platform used at most Bioinformatics laboratories in developing countries and, for this reason, this is the target platform used in this paper.…”

Comparison of HPC Architectures for Computing All-Pairs Shortest Paths. Intel Xeon Phi KNL vs NVIDIA Pascal

“…Because each accelerator has its advantages and disadvantages for certain classes of problems [22,3,17], selecting the best option for a given application is key when searching for maximum performance. To provide some guidelines for such selection, this article presents a comparative analysis between two different HPC architectures (Intel Xeon Phi KNL vs. NVIDIA Pascal).…”