Scalability of Self-organizing Maps on a GPU cluster using OpenCL and CUDA
Sabine McConnell, Robert Sturgeon, Gregory W. Henry, Andrew J. Mayne, Richard T. Hurley
Abstract
Open-access reader
Sabine McConnell, Robert Sturgeon, Gregory W. Henry, Andrew J. Mayne, Richard T. Hurley
Abstract
Open-access reader
We evaluate a novel implementation of a Self-Organizing Map (SOM) on a Graphics Processing Unit (GPU) cluster. Using various combinations of OpenCL, CUDA, and two different graphics cards, we demonstrate the scalability of the SOM implementation on one to eight GPUs. Results indicate that while the algorithm scales well with the number of training samples and the map size, the benefits from using the data-parallel approaches offered by the GPU are severely limited when combined with the Message Passing Interface (MPI) in this setting, and comparable to speedups of GPU-based implementations as compared to optimized sequential code. Speedups achieved range from 3 to 32, for various map and training data sizes. We also observed a performance penalty for the OpenCL implementation as compared to CUDA.
OpenAlex reports 33 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
We evaluate a novel implementation of a Self-Organizing Map (SOM) on a Graphics Processing Unit (GPU) cluster. Using various combinations of OpenCL, CUDA, and two different graphics cards, we demonstrate the scalability of the SOM implementation on one to eight GPUs. Results indicate that while the algorithm scales well with the number of training samples and the map size, the benefits from using the data-parallel approaches offered by the GPU are severely limited when combined with the Message Passing Interface (MPI) in this setting, and comparable to speedups of GPU-based implementations as compared to optimized sequential code. Speedups achieved range from 3 to 32, for various map and training data sizes. We also observed a performance penalty for the OpenCL implementation as compared to CUDA.
Key concepts: CUDA, Computer science, GPU cluster, Scalability, Parallel computing, Graphics processing unit, General-purpose computing on graphics processing units, Graphics