Precision Health Initiative Colloquium
Speaker: Ariful Azad, Lawrence Berkeley National Laboratory
Where: Luddy Hall, Rm. 1106 (Dorsey Learning Hall)
Topic: Massively-Parallel Graph Analytics for Exascale Applications
Abstract: A graph succinctly captures complex relationships among entities arisen in scientific and business datasets such as interacting proteins, cosmological particles, or friends on social networks. Recently, many science applications started generating massive graphs with billions of nodes and trillions of connections. To extract information from these massive graphs, we need massively-parallel algorithms capable of utilizing current and future computing platforms. In this talk, I will discuss a family of massively-parallel graph algorithms that scales to hundreds of thousands of processors. These algorithms enable discoveries in several high-impact scientific domains including genomics, neuroscience, immunology, scientific computing, quantum computing and machine learning. For example, using large supercomputers, a parallel Markov clustering algorithm is able to cluster large-scale biological networks created from genomic and metagenomics data, enabling unprecedented discoveries in network biology. I will also discuss algorithmic foundations for identifying phenotypic signatures of Acute Myeloid Leukemia, mapping brain segments from functional MRI data and solving sparse linear systems of equations. I will finish with a sketch of a data analytics framework that can exploit millions of processors to be available in the upcoming exascale systems. The proposed framework can stimulate cross-disciplinary research by solving high-impact science and business applications with assured high performance at a lower development cost.
Bio: Ariful Azad is Research Scientist in the Computational Research Division at the Lawrence Berkeley National Laboratory (LBNL). He obtained is Ph.D. from the department of Computer Science at Purdue University and BS from Bangladesh University of Engineering and Technology. His research interests are in parallel graph algorithms, high performance computing, data-intensive computing, and Bioinformatics applications.