vix.ing · top · new · best · stats · spec

SparkCL: A Unified Programming Framework for Accelerators on\n Heterogeneous Clusters

2015/05/05 by Oren Segal, Philip Colangelo, Segal, Oren +7 · 1 citation
Computer Science · Decision Sciences · #Advanced Data Storage Technologies #Distributed #Distributed and Parallel Computing Systems #FOS: Computer and information sciences #Parallel #Parallel Computing and Optimization Techniques #Scientific Computing and Data Management #and Cluster Computing (cs.DC)

paper · pdf · doi:10.48550/arxiv.1505.01120

openalex publication_date 2015/05/05 · openalex created_date 2022/09/01 · openalex updated_date 2026/07/28

Abstract

We introduce SparkCL, an open source unified programming framework based on\nJava, OpenCL and the Apache Spark framework. The motivation behind this work is\nto bring unconventional compute cores such as FPGAs/GPUs/APUs/DSPs and future\ncore types into mainstream programming use. The framework allows equal\ntreatment of different computing devices under the Spark framework and\nintroduces the ability to offload computations to acceleration devices. The new\nframework is seamlessly integrated into the standard Spark framework via a\nJava-OpenCL device programming layer which is based on Aparapi and a Spark\nprogramming layer that includes new kernel function types and modified Spark\ntransformations and actions. The framework allows a single code base to target\nany type of compute core that supports OpenCL and easy integration of new core\ntypes into a Spark cluster.\n

Cited by

Related