Advertisement

Researchers have designed a computing resource, called ConFlux, to enable HPC clusters to communicate seamlessly and at interactive speeds with data-intensive operations. Hosted at U-M, the project establishes a hardware and software ecosystem to enable large scale data-driven modeling of complex physical problems, such as the performance of an aircraft engine which consists of trillions of molecular interactions.San Jose, CA — The University of Michigan (U-M) announced it has selected IBM to develop and deliver “data-centric” supercomputing systems designed to increase the pace of scientific discovery in fields as diverse as aircraft and rocket engine design, cardiovascular disease treatment, materials physics, climate modeling and cosmology.

Traditionally, scientific computations have been performed on high performance computing (HPC) infrastructure while modern data parallel architectures have mostly focused on web analytics and business intelligence applications. Systems that enable HPC applications for physics to interact in real time with big data to improve quantitative predictability have not yet been developed. IBM’s systems use a data-centric approach, integrating massive datasets seamlessly with HPC computing power resulting in new predictive simulation techniques that will expand the limits of scientific knowledge.

Working with IBM, U-M researchers have designed a computing resource, called ConFlux, to enable HPC clusters to communicate seamlessly and at interactive speeds with data-intensive operations. Hosted at U-M, the project establishes a hardware and software ecosystem to enable large scale data-driven modeling of complex physical problems, such as the performance of an aircraft engine which consists of trillions of molecular interactions. ConFlux will produce advances in predictive modeling in several fields of computational science, and is funded by a grant from the National Science Foundation.

“There is a pressing need for data-driven predictive modeling to help re-envision traditional computing models in our pursuit to bring forth groundbreaking research,” said Karthik Duraisamy, Assistant Professor, Department of Aerospace Engineering and Director, Center for Data-driven Computational Physics, U-M. “The recent acceleration in computational power and measurement resolution has made possible the availability of extreme scale simulations and data sets. ConFlux allows us to bring together large scale scientific computing and machine learning for the first time to accomplish research that was previously impossible.”

ConFlux meshes well with IBM’s focus on data-centric computing systems.

“Scientific research is now at the crossroads of big data and high performance computing,” said Sumit Gupta, Vice President, High Performance Computing and Data Analytics, IBM. “The explosion of data requires systems and infrastructures based on POWER8 plus accelerators that can both stream and manage the data and quickly synthesize and make sense of data to enable faster insights.”

"U-M grasped the significance of IBM's shift to data-centric systems during our first discussion,” said Michael J. Henesey, Vice President Business Development, Data Centric Systems and Innovation Centers. “They were enthusiastic about the application of this architecture to problems that are essential to the University and to the country. We will stay close to U-M to help inform our future system designs."

IBM OpenPOWER systems drive powerful computing

Progress in a wide spectrum of fields ranging from medicine to transportation relies critically on the ability to gather, store, search and analyze big data and construct truly predictive models of complex, multi-scale systems.

Advanced technologies like data-centric computing systems are at the forefront of tackling these big data challenges and advancing the pace of innovation. By moving computing power to where the data resides, organizations of all sizes can maximize performance and minimize latency in their systems, enabling them to gain deeper insights from research. These data-centric solutions are accelerated through open innovation and IBM’s work with other members of the OpenPOWER Foundation.

The incorporation of OpenPOWER technologies into a modular integrated system will enable U-M to configure the systems for their specific needs. ConFlux incorporates IBM Power Systems LC servers, which were designed based on technologies and development efforts contributed by OpenPOWER Foundation members including Mellanox, NVIDIA, Tyan and Wistron.

Additional data-centric solutions U-M is using include IBM Elastic Storage Server, IBM Spectrum Scale software (scale-out, parallel access network attached storage), and IBM Platform Computing software.

In an internal comparison test conducted by U-M, the POWER8 system significantly outperformed a competing architecture by providing low latency networks and a novel architecture that allows for the integrated use of central and graphics processing units.

Data-centric, cognitive approach advances research at U-M

One of the first projects U-M will undertake with its advanced supercomputing system is working with NASA to simulate turbulence around aircraft and rocket engines through cognitive techniques. Large amounts of data from wind tunnel experiments and simulations are combined to build computing models that are used to predict the aerodynamics around new configurations, such as an aircraft wing or engine. With ConFlux, U-M can more accurately model and study turbulence, helping to speed development of more efficient airplane designs. It will also improve weather forecasting, climate science and other fields that involve the flow of liquids or gases.

U-M is also studying cardiovascular disease for the National Institutes of Health. By combining noninvasive imaging such as MRI and CT scan results with a physical model of blood flow, U-M hopes to help doctors estimate artery stiffness within an hour of a scan, serving as an early predictor of diseases such as hypertension.

Studies are also planned to better understand climate science such as how clouds interact with atmospheric circulation, the origins of the universe and stellar evolution, and predictions of the behavior of biologically inspired materials.

"The ConFlux project aligns with the University of Michigan's comprehensive strategy of investment in research computing and data science across disciplines," said Eric Michielssen, U-M's Associate Vice President for Research Computing. “For example, our $100 million Data Science Initiative is advancing faculty driven research in engineering and the social and health sciences by building connections between the worlds of big data and HPC. ConFlux epitomizes this forward-looking vision.”

Advertisement
Advertisement