Gravitational search algorithm using CUDA: a case study in high-performance metaheuristics
Amirreza Zarrabi, Khairulmizam Samsudin, Ettikan K. Karuppiah
The Journal of Supercomputing, December 2014
Many scientific and technical problems with massive computation requirements could benefit from the graphics processing units (GPUs) using compute unified device architecture (CUDA). Gravitational search algorithm (GSA) is a population-based metaheuristic which can be effectively implemented on GPU to reduce the execution time. Nonetheless, the performance improvement depends strongly on the process used to adapt the algorithm into CUDA environment. In this paper, we discuss possible approaches to parallelize GSA on graphics hardware using CUDA. An in-depth study of the computation efficiency of parallel algorithms and capability to effectively exploit the architecture of GPU is performed. Additionally, a comparative study of parallel and sequential GSA was carried out on a set of standard benchmark optimization functions. The results show a significant speedup while maintaining results quality which re-emphasizes the utility of CUDA-based implementation for complex and computationally intensive parallel applications.
Galactica : A GPU Parallelized Database Accelerator
KK Yong, Ettikan K. Karuppiah, Simon See
The Third ASE International Conference on Big Data Science and Computing, Beijing, China; 08/2014
The amount of business data generated and collected is increasing exponentially every year. A Graphics Processing Unit (GPU) is not used for only optimization of image filtering and video processing, but is also widely adopted for accelerating big data analytics for scientific, engineering, and enterprise applications. However, there are studies pointing out that using GPU as a general-purpose computing device has limitations. In order to exploit current GPU computing capabilities for database operations, we have to take into consideration the characteristics of the GPU and how it can cooperate with the CPU. In this paper, we proposed and implemented a GPU database accelerator, which named Galactica. The experiments result shows proposed GPU database accelerator has outperformed traditional database system. In addition, the Galactica’s performance is comparable with a seven nodes distributed Hadoop system. Our results indicate that the GPU is an effective and energy efficient co-processor for executing database operations.
Gravitational Search Algorithm using CUDA
Amirreza Zarrabi, Ettikan K. Karuppiah, Yong Keh Kok, Ngo Chuan Hai, Simon See
IEEE Parallel and Distributed Computing, Applications and Technologies (PDCAT 2014); 12/2014
Many scientific and technical problems with massive computation requirements could benefit from the Graphics Processing Units (GPUs) using Compute Unified Device Architecture (CUDA) for high speed processing. Gravitational Search Algorithm (GSA) is a population-based metaheuristic algorithm that can be effectively implemented on GPU to reduce the execution time. In this paper we discuss possible approaches to parallelize GSA on graphics hardware using CUDA. An indepth study of the computation efficiency of parallel algorithms and capability to effectively exploit the architecture of GPU is performed. Additionally, a comparative study of parallel and sequential GSA was carried out on a set of standard benchmark optimization functions. The results show a significant speedup that re-emphasizes the utility of CUDA based implementation for complex and computationally intensive parallel applications.
Multi Keyword Range Search in GPU and MIC: A Comparison Study
Amirul Abdullah, Kek Kok Yong, Ettikan K. Karuppiah, Poh Kit Chong
IEEE Conference on Open Systems (ICOS 2014), Kuala Lumpur, Malaysia; 10/2014
Data, both structured and unstructured, is increasing exponentially daily. This valuable data is important to businesses, society, and other organisations in order to compute more accurate analysis, and eventually, make better judgement. In order to handle huge data, many have turned to co-processors like GPUs or Intel MIC to further accelerate their computation. In this study, we present performance and evaluation comparison of GPU and MIC by implementing Multi Text Keyword Search algorithms from our prior work into MIC and GPU. We use NVIDIA K20c and NVIDIA K40 for our GPUs and Intel® Xeon Phi™ 5100 for the MIC. In our experiments and from our observation we found out K20c and K40 outperformed MIC for this particular algorithm.
Open Platform for Advanced Analytics, Meeting Data Convergence Challenge
Ettikan Karuppiah, KK Yong, Fazli Mat Nor
United Nations Global e-Government Forum, Astana, Kazakhstan; 10/2014
Data size for interest specific analytics is growing in exponential manner turning data into as ‘real currency’. Thus, MIMOS believes in an open platform/middleware framework that enables automatic distribution of various data and processing across heterogeneous computing resources for structured and unstructured data analytics. Thus, we propose an open platform/middleware framework for data analytics that provides mechanisms for automatic data segmentation, distribution, execution, information retrieval across multiple processors (CPU & GPU) and machines, a modular design for easy addition of new GPU kernels at both analytic and processing layer, and information presentation. Our results show proposed middleware framework provides alternative and cheaper HPC solution to users. For example, data-cleansing algorithms on GPU shows a speed up of over two orders of magnitude compared to the same operation done in MySQL on a multi-core machine. Our framework is also capable of processing more than 120 million of health data within 11 seconds.
Big Data Analytics Using Accelerator for HPC
KK Yong, Boon Keong Seah, Ettikan K. Karuppiah
HPC Advisory Council Singapore Conference 2014, Singapore, Singapore; 10/2014
We are living in a data inundation era. This has brought a big impact and attracted diversify thoughtfulness from both technological experts and the public. We are experimenting the advancement of HPC technologies, and also designing the refined parallel algorithms. Thus, it leads us to transform an adversity into fortune.