Michigan Institute for Data Science (MIDAS) announces second round of Challenge Thrust competition: Social Science & Health Science

By | General Interest, News | No Comments

The Michigan Institute for Data Science (MIDAS) is pleased to announce the second competition for MIDAS Challenge Thrust awards. These awards are intended to stimulate research in key areas identified at our inaugural symposium last fall, and will lay the foundation for future funding from government, private foundations, or industry.

Requests for Proposals (RFPs) are available for awards in Data Science for Social Science and Health Science. Up to two projects will be funded at a level of approximately $1.25 million each in both of these Challenge Thrust areas.

White papers describing project goals and teaming arrangements are due May 16, 2016, and full proposals are due September 12, 2016. Awards will be announced on October 10, 2016. 

Successful research projects will cut across disciplines, have the potential for disruptive impact in the field, and hold promise for advancing the methodological foundations of data science. Interested researchers can learn more about these two MIDAS Challenge Thrust areas and connect with potential collaborators at two upcoming town hall meetings.

Upcoming Town Halls

Health Science

  • Wednesday, March 30, 2016, 5-7 p.m., Henderson Room, Michigan League

Social Science

  • Thursday, March 31, 2016, 5-7 p.m., Institute for Social Research

RFPs for the MIDAS Challenge Thrust awards in the Learning Analytics and Transportation areas were released in fall 2015. Awards in those areas will be announced this spring.

For more information, email midas-rfp@umich.edu, or visit our RFP page at midas.umich.edu/rfp.

New nodes added to increase capacity of Flux HPC cluster

By | General Interest, News | No Comments

As part of regular upgrades to our high performance computing resources, Advanced Research Computing – Technology Services (ARC-TS) has increased the capacity of the Flux HPC cluster.

These newest (seventh generation) nodes include:

  • 167 Standard Flux nodes with 24 cores and 128 GB RAM each
  • 4 Large Memory Flux nodes with 56 cores and 1.5 TB RAM each
  • Connected via EDR (100 Gbps) Infiniband
  • Adds a total of 4,232 cores and 27 TB RAM to Flux.

For complete descriptions of the various Flux configurations, see our HPC Resources page.

The additional resources allow Flux to handle more jobs simultaneously and increases the number of cores and amount of memory on the largest nodes.

Jobs can request the new cores by using the standard PBS options (http://arc-ts.umich.edu/software/torque/) and/or requesting the node property “haswell”. Please note that the new nodes are likely to be popular and jobs that request the new nodes may wait longer to start than jobs with less-specific resource requirements.

ARC Director Sharon Broude Geva to deliver keynote address at PRACEdays16 in Prague

By | General Interest, News | No Comments

Sharon Broude Geva, Director of Advanced Research Computing (ARC), will deliver a keynote address at PRACEdays16 in Prague, Czech Republic in May.

PRACE, the Partnership for Advanced Computing in Europe, has 25 member countries; its mission is to enable high impact scientific discovery and engineering research and development across all disciplines to enhance European competitiveness for the benefit of society. PRACE seeks to realize this mission by offering world class computing and data management resources and services through a peer review process.

Geva’s talk is titled “Beyond Hardware: Scaling Up the Approach to Advanced Research Computing.”

Abstract (excerpt): With the surge in disciplines and research communities using advanced research computing, the scope of required support for High-Performance Computing has expanded. It is no longer enough to provide hardware and the people to build it and keep it running. Facilitating the use of HPC and other technology by users with little experience in coding; creating academic programs for undergraduate and graduate students; training students, faculty, and staff to use new technologies; and providing consulting, programming, and visualization services are but a few of the aspects that Universities are called on to address.

For more information: https://events.prace-ri.eu/event/488/

 

REMINDER: Info Sessions: Graduate programs in computational and data science — Feb. 22, 23

By | Educational, Events | No Comments

Learn about graduate programs that will prepare you for success in computationally intensive fields, and enjoy some pizza. Presentations will describe the following programs:

  • The Ph.D. in Scientific Computing is open to all Ph.D. students who will make extensive use of large-scale computation, computational methods, or algorithms for advanced computer architectures in their studies. It is a joint degree program, with students earning a Ph.D. from their current departments, “… and Scientific Computing” — for example, “Ph.D. in Aerospace Engineering and Scientific Computing.”
  • The Graduate Certificate in Computational Discovery and Engineering trains graduate students in computationally intensive research so they can excel in interdisciplinary HPC-focused research and product development environments. The certificate is open to all students currently pursuing Master’s or Ph.D. degrees at the University of Michigan.
  • The Graduate Certificate in Data Science is focused on developing core proficiencies in data analytics: 1) Modeling — Understanding of core data science principles, assumptions and applications; 2) Technology — Knowledge of basic protocols for data management, processing, computation, information extraction, and visualization; 3) Practice — Hands-on experience with real data, modeling tools, and technology resources.

Monday, February 22, 5-6 p.m., Room 2001, LS&A Building, 500 State St. 

Tuesday, February 23, 5-6 p.m., EECS 1200, 1301 Beal Ave. 

Presenters:

  • Krishna Garikipati, Professor of Mechanical Engineering and Mathematics, and Associate Director for Research, Michigan Institute for Computational Discovery and Engineering.
  • Ivo Dinov, Associate Professor of Computational Medicine and Bioinformatics, and Human Behavior and Biological Sciences.
  • Ken Powell, Arthur F. Thurnau Professor of Aerospace Engineering.

There will be time for questions and discussion.

HIPAA-aligned HPC cluster “Armis” now available

By | General Interest, News | No Comments

U-M researchers who manage and analyze large volumes of data protected under the Health Insurance Portability and Accountability Act (HIPAA) now have a new secure computing tool at their disposal: the HIPAA-aligned “Armis” high performance computing cluster.

Armis operates in tandem with the HIPAA-aligned Turbo Research Storage service, which has been available since Fall 2015, to provide a comprehensive computing and storage platform to unleash the power of high performance computing for research involving HIPAA-protected data.

“Armis will allow researchers in fields ranging from medicine and public health to the biological and social sciences to draw on powerful high performance computing tools while adhering to the requirements of HIPAA,” said Eric Michielssen, Associate Vice President – Advanced Research Computing. “It will open up new opportunities for faculty and students and further boost interdisciplinary research engendered by the Data Science Initiative.”

Armis is available to all researchers on campus, and operated by Advanced Research Computing – Technology Services (ARC-TS). Armis is accessed and used in essentially the same way as the Flux HPC cluster. Users of Armis have access to the same software as Flux users, including statistical packages like Stata, R, and SAS.

“Armis will prove immensely useful to healthcare researchers working with sensitive patient information. As we try to fulfill the promise of big data in medicine and turn precision medicine from an idea into a reality, we need to analyze large volumes of diverse data types. Turning billions of clinical observations, genomic data elements, and unstructured data into something that impacts real patient care requires the high performance computing power of Armis,” says Sachin Kheterpal, MD, MBA, Senior Director of the University of Michigan Medical School Research Data Warehouse.

Use of Armis has certain advantages even for researchers using data that do not directly fall under HIPAA regulations. Researchers using other forms of sensitive data requiring stricter controls and security practices (for example, through agreements with data providers or institutional review boards) might find that Armis may meet their needs.

For more information on Armis:

Please email arc-contact@umich.edu with any questions.

 

MICDE Seminar: Jim Belak, Lawrence Livermore National Lab, “Bridging Scales within Exascale Materials Co-design Center” — Feb. 5

By | Educational, Events | No Comments

As part of the MICDE Seminar Series, Jim Belak of the Materials Science Division at Lawrence Livermore National Laboratory will speak on campus Friday, Feb. 5.

Abstract: The advent of Advanced / Additive Manufacturing and the Materials Genome Initiative has placed significant emphasis on accelerating the qualification of new materials for use in real applications. Within these workflows lies both the engineering scale qualification through building and testing components at scale and full-scale modeling with integrated continuum computer codes and the materials scale qualification through revolutionary methods to non-destructively measure microstructure (3DXRD) and physics specific experiments coupled with meso-scale mechanics simulations of the same physics specific experiment using the same microstructure. This ICME process is one of the use cases that drives the Exascale Materials Co-design Center (ExMatEx). The goal of the Co-design Center is very analogous to the acceleration of new materials deployment within the MGI, rather co-design accelerates the deploying of laboratory concepts for future computer components to enable a productive exascale computer system. To enable better meso-scale understanding in the continuum models, ExMatEx is creating a direct coupling between the continuum integrated code and direct numerical simulation of the meso-scale phenomena. Here we review the ExMatEx project, and its use cases.

Bio: Jim Belak is an Applied Scientist in the Materials Science Division at Lawrence Livermore National Laboratory. His career has centered around the application of High Performance Computing to equilibrium and non-equilibrium problems in Condensed Matter Physics, including: order-disorder phase transition in solids; indentation, metal cutting and tribology of interfaces; shock propagation and spallation fracture; structure and dynamics of grain boundaries and defects in solids; and kinetics of phase evolution in extreme environments. These applications have required the development of new algorithms and application codes for emerging high performance parallel computers and the use of novel x-ray synchrotron techniques (3D x-ray tomography and small-angle x-ray scattering) to guide and validate the simulations. Currently, Jim co-leads the Exascale Co-design Center for Materials in Extreme Environments (ExMatEx).

MIDAS Seminar: Kevin Ward, Executive Director, Michigan Center for Integrative Research in Critical Care, “Data in Motion Phenotyping” — Feb. 5

By | Educational, Events | No Comments

MIDAS is proud to announce that Kevin Ward, Professor of Emergency Medicine and Executive Director of the Michigan Center for Integrative Research in Critical Care, and Fast Forward Medical Innovation, will speak at 4 p.m., Feb. 5, 2016, in the Michigan League Ballroom, as part of the MIDAS Seminar Series. All seminars are streamed live (sign in as Guest).

CSCAR Data Science Skills Series adds session on Pandas case studies — Feb. 17

By | Educational, Events | No Comments

CSCAR is offering a series of workshops on data science skills using Python. The workshops will be held in the Earl Lewis room in the Rackham building. All workshops will take place on Wednesday afternoons from 3:30-5.

No registration is necessary and there is no fee to attend. Please note: A new section has been added on Pandas case studies.

Schedule:

  • January 27: Data management with Pandas
  • February 10: Graphics and data visualization with Matplotlib and Bokeh
  • NEW: February 17: Pandas case studies (CMS data analyses)
  • February 24: Basic statistical analysis with Statsmodels
  • March 9: Sklearn for predictive analysis and data exploration
  • March 23: Advanced regression analysis (GEE, mixed models and multiple imputation) with Statsmodels
  • April 6: Survival analysis with Statsmodels

Additional workshops will be scheduled on the following topics, dates to be announced:

  • Geospatial analysis
  • Building and accessing databases
  • MPI, parallel, and distributed computing

Class material will be posted on the series website.

Save the Date: MICDE Symposium, April 7

By | Educational, Events | No Comments

The Michigan Institute for Computational Discovery and Engineering (MICDE) Annual Symposium will take place April 7 in the Rackham Building on U-M’s Central Campus.

Titled “Towards Tomorrow’s Computational Science,” the symposium will feature an outstanding group of speakers, including the director of NSF’s Division of Advanced Cyberinfrastructure, Irene Qualters; 2016 SIAM awardee Linda Petzold (UCSB); AMS/SIAM Norbert Wiener Prize winner James Sethian (Berkeley); and MathWorks co-founder and Matlab author Cleve Moler.

The symposium will also include a poster session highlighting outstanding computational work from U-M researchers and students. To participate in the poster session, contact Mariana Carrasco-Teja.

More information, including a detailed agenda, will be posted on the MICDE website as it becomes available.

Study seeks input on U-M HPC support

By | General Interest | No Comments

Are you a researcher utilizing High Performance Computing (HPC) resources? Have you found these resources confusing or challenging to use? Have your support concerns, if any, been properly and promptly addressed?

A group of students at the School of Information would like to talk to you about your experiences.  HPC support staff will use recommendations from these interviews to enhance their service. Interviews will take place once or twice during the semester for approximately one hour each.  Please contact us at umsi-hpc-support-study@umich.edu  if you would like to participate.