Energy-efficient FPGA Implementation of the k-Nearest Neighbors Algorithm Using OpenCL

Abstract

Abstract-ModernSoCs are getting increasingly heterogeneous with a combination of multi-core architectures and hardware accelerators to speed up the execution of computeintensive tasks at considerably lower power consumption. Modern FPGAs, due to their reasonable execution speed and comparatively lower power consumption, are strong competitors to the traditional GPU based accelerators. High-level Synthesis (HLS) simplifies FPGA programming by allowing designers to program FPGAs in several high-level languages e.g. C/C++, OpenCL and SystemC.This work focuses on using an HLS based methodology to implement a widely used classification algorithm i.e. k-nearest neighbor on an FPGA based platform directly from its OpenCL code. Multiple fairly different implementations of the algorithm are considered and their performance on FPGA and GPU is compared. It is concluded that the FPGA generally proves to be more power efficient as compared to the GPU. Furthermore, using an FPGA-specific OpenCL coding style and providing appropriate HLS directives can yield an FPGA implementation comparable to a GPU also in terms of execution time.

Optimizing OpenCL Code for Performance on FPGA: k-Means Case Study With Integer Data Sets

“…It is difficult to draw a straightforward comparison due to the diversity of baselines, data sets, and target devices. We find that the approaches most similar to our own, considering the baseline, approach, and platform, are Tang et al [29] and Muslim et al [30]. Regarding Tang et al [29], the k-means algorithm was also implemented via OpenCL on a comparable FPGA.…”

“…Muslim et al study the kNN algorithm [30]. The algorithm calculates which k nearest points in a training set are closest to each point n of the input data set.…”

High Performance and Low Power Monte Carlo Methods to Option Pricing Models via High Level Design and Synthesis

“…A comparison between GPU and FPGA has also been performed for several other algorithms, such as the k-Nearest Neighbor in [7].…”

“…The numerical solution for (5) is also obtained by Euler discretization [10] with full truncation scheme avoiding negative values under the square root [8]. The final solutions are shown in (6) and (7).…”

“…GPUs or FPGAs). The basic architecture of such a heterogeneous framework has also been depicted and implemented in several other literatures such as in [7].…”