Graduate Seminars

back Back to all Graduate Seminars

CS Graduate Seminar: Matrix Algebra Framework for Portable, Scalable and Efficient Query Engines for RDF Graphs

Start Date: December 3, 2018
End Date: December 3, 2018

By PhD candidate Fuad Jamour (KAUST)
 
Existing query engines for RDF data are typically designed for specific hardware architectures, mainly CPUs. Porting such engines to a different architecture (e.g., GPUs) entails almost complete redesign. We propose MAGiQ; a framework for implementing SPARQL query engines that are portable on various hardware architectures, scalable over thousands of compute nodes, and efficient for very large RDF datasets. MAGiQ represents the RDF graph as a sparse matrix and defines a domain-specific language of algebraic operations. SPARQL queries are translated into matrix algebra programs that are oblivious to the underlying computing infrastructure. Then, existing matrix algebra libraries, optimized for each particular architecture, are called to execute the program and handle all related issues including data partitioning, parallel processing, communication and load balancing. We present three case studies of back-end matrix algebra libraries, namely SuiteSparse, CombBLAS and Matlab, and demonstrate that MAGiQ can be ported effortlessly on a variety of architectures, such as Intel CPUs, NVIDIA GPUs and a Cray XC40 supercomputer. We show that MAGiQ's performance is comparable to or better than existing specialized RDF systems. We also show that MAGiQ handles datasets with up to 128 billion triples and scales to 4096 cores.
 
Bio: Fuad Jamour is a PhD candidate in the computer science program at KAUST. He holds a BSc from the University of Jordan and an MS from KAUST. His research interests include scalable graph processing, distributed systems, and big data management. His work appeared in several top-tier journals and conferences such as IEEE TPDS, VLDB, and SC.
 

More Information:

For more info contact: Mr. Fuad Jamour : email: fuad.jamour@kaust.edu.sa
 
Date: Monday 3rd Dec 2018
Time:12:00 PM - 01:00 PM
Location: Engineering Science Hall (bldg.9), Level 2, Hall 1
Refreshments: Light Lunch will be available at 11:45 am