Intel® Parallel Computing Center at Indiana University

Published: 07/31/2015, Last Updated: 07/19/2018

Principal Investigators:

Judy Qiu

Judy Qiu is an associate professor of Intelligent Systems Engineering at Indiana University. Her area of research is in data-intensive computing at the intersection of Cloud and high performance computing (HPC) multicore technologies. This includes a specialization on programming models that support iterative computation, ranging from storage to analysis which can scalably execute data intensive applications. Her research has been funded by NSF, NIH, Intel, Microsoft*, Google*, and Indiana University. She is the recipient of a NSF CAREER Award, Indiana University Trustees Award for Teaching Excellence, and Indiana University Outstanding Junior Faculty Award.

Description:

The Indiana University Intel® Parallel Processing Center (Intel® PCC) is a multi-component interdisciplinary center. The focus of the center is grand challenges in high performance simulation and data analytics with innovative applications, and software development using the Intel® architecture. Issues of programming productivity and performance portability will be studied.

High Performance Data analysis plays an important role in both data-driven scientific discovery and commercial services with research on novel parallel systems supporting data analytics and machine learning. The earlier research has introduced a novel HPC-Cloud convergence framework named Harp-DAAL and demonstrated that the combination of Big Data and HPC techniques can simultaneously achieve productivity and performance. Harp is a communication library (as a plugin to Apache Hadoop*) that orchestrates efficient node synchronization with a MapCollective computing model optimized for big data. Harp uses Intel® Data Analytics Accelerator Library (Intel® DAAL), for its highly optimized kernels on Intel® Xeon and Intel® Xeon Phi™. This way the high-level API of big data tools can be combined with intra-node fine-grained parallelism that is optimized for HPC platforms. Harp-DAAL shows how simulations and big data can use common programming environments with a runtime based on a rich set of collectives and libraries.

The project will select a subset of machine learning and data analysis algorithms from state-of-the-art libraries and tools and compare them on different platforms (including Xeon, Xeon Phi and other accelerators). We will benchmark the Harp-DAAL machine learning library and compare it with other Machine Learning and Deep Learning frameworks. We will produce a high-quality Machine Learning library with optimal performance, which compares with routines from Apache Mahout, MLlib, TensorFlow* and other libraries such as those in R and Python*.

Publications:

Related Websites:

http://ipcc.soic.iu.edu/

Product and Performance Information

1

Intel's compilers may or may not optimize to the same degree for non-Intel microprocessors for optimizations that are not unique to Intel microprocessors. These optimizations include SSE2, SSE3, and SSSE3 instruction sets and other optimizations. Intel does not guarantee the availability, functionality, or effectiveness of any optimization on microprocessors not manufactured by Intel. Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors. Certain optimizations not specific to Intel microarchitecture are reserved for Intel microprocessors. Please refer to the applicable product User and Reference Guides for more information regarding the specific instruction sets covered by this notice.

Notice revision #20110804