Skip to main content

Table 12 GPUs contribution in acceleration and energy consumed versus a 2017 FPGA implementation using OpenCL

From: DNA sequences alignment in multi-GPUs: acceleration and energy payoff

Platform

FPGA

GTX 980

GTX 980

GTX 980

Titan Pascal

Implementation

OpenCL

SW# (CUDA)

CUDAlign 3.0

CUDAlign 4.0

CUDAlign 4.0

Year

2017

2013

2014

2016

2017

Performance

Baseline

+79.3%

+125.2%

+213.5%

+669.2%

Power efficiency

Baseline

- 72.6%

- 66.0%

- 23.3%

- 1.3%