FlinkCL: An OpenCL-Based In-Memory Computing Architecture on Heterogeneous CPU-GPU Clusters for Big Data
- 23 May 2018
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 67 (12) , 1765-1779
Abstract
Research on in-memory big data management and processing has been prompted by the increase in main memory capacity and the explosion in big data. By offering an efficient in-memory distributed execution model, existing in-memory cluster computing platforms such as Flink and Spark have been proven to be outstanding for processing big data. This paper proposes FlinkCL, an in-memory computing architecture on heterogeneous CPU-GPU clusters based on OpenCL that enables Flink to utilize GPU's massive parallel processing ability. Our proposed architecture utilizes four techniques: a heterogeneous distributed abstract model (HDST), a Just-In-Time (JIT) compiling schema, a hierarchical partial reduction (HPR) and a heterogeneous task management strategy. Using FlinkCL, programmers only need to write Java code with simple interfaces. The Java code can be compiled to OpenCL kernels and executed on CPUs and GPUs automatically. In the HDST, a novel memory mapping scheme is proposed to avoid serialization or deserialization between Java Virtual Machine (JVM) objects and OpenCL structs. We have comprehensively evaluated FlinkCL with a set of representative workloads to show its effectiveness. Our results show that FlinkCL improve the performance by up to $11\times$ for some computationally heavy algorithms and maintains minor performance improvements for a I/O bound algorithm.
Keywords
Funding Information
- National Outstanding Youth Science Program of National Natural Science Foundation of China (61625202)
- International (Regional) Cooperation and Exchange Program of National Natural Science Foundation of China (61661146006)
- Singapore-China NRF-NSFC (NRF2016NRF-NSFC001-111)
- National Natural Science Foundation of China (61370095, 61472124, 61662090, 61602350)
- Key Technology Research and Development Programs of Guangdong Province (2015B010108006)
- Outstanding Graduate Student Innovation Fund Program of Collaborative Innovation Center of High Performance Computing
- China Scholarship Council (CSC) (201706130121)
This publication has 24 references indexed in Scilit:
- HadoopCL2: Motivating the Design of a Distributed, Heterogeneous Programming System With Machine-Learning ApplicationsIEEE Transactions on Parallel and Distributed Systems, 2015
- Data Partitioning on Multicore and Multi-GPU Platforms Using Functional Performance ModelsIEEE Transactions on Computers, 2014
- Performance Optimization Using Partitioned SpMV on GPUs and Multicore CPUsIEEE Transactions on Computers, 2014
- Scaling MapReduce Vertically and Horizontally2014
- MRONLINE2014
- BigKernel -- High Performance CPU-GPU Communication Pipelining for Big Data-Style Applications2014
- Processing MPI Derived Datatypes on Noncontiguous GPU-Resident DataIEEE Transactions on Parallel and Distributed Systems, 2013
- Optimizing MapReduce for GPUs with effective shared memory usage2012
- OpenMP: an industry standard API for shared-memory programmingIEEE Computational Science and Engineering, 1998
- A high-performance, portable implementation of the MPI message passing interface standardParallel Computing, 1996