U-M professor Quentin Stout, a veteran of all 28 Supercomputing conferences, reflects on SC through the years

By | Uncategorized | No Comments
StoutJablonowski

Quentin Stout and Christian Jablonowski teaching the Parallel Computing 101 tutorial at SC07.

Quentin Stout, University of Michigan Professor of Computer Science and Engineering (CSE) and Climate and Space Sciences and Engineering (CLaSP), has attended all 28 of the Supercomputing conferences since the event begin in 1988. Stout is one of less than 20 so-called “SC Perennials” to have attended every one. He, along with Christiane Jablonowski, associate professor in CLaSP, have taught the Introduction to Parallel Computing tutorial at the conference for many years and are teaching it again this year. Stout, who has been at U-M since 1984, recently answered some questions about the evolution of the field of computer science and the area of supercomputing over the decades.

Question: What was the first SC conference like, and how has it changed over the years?

Stout: The first conference, in 1988, had about 1,500 people, compared to the over 10,000 now. Its focus was on supercomputing and the large centers at DOE, NASA, NSF, etc., along with the companies that were making these systems. There were also some researchers from academia and a few industrial users. The largest supercomputer user, NSA, had people at the conference, but they didn’t have a booth and their badge listed “Fort Meade” as their affiliation.

Over the years it has greatly broadened its scope to have a much broader international focus and more participation by universities, cluster vendors of all sizes, networking, storage, commercial software, educational efforts, etc. …

Originally I went to learn more about the field, meet people, see what the emerging areas were, and learn about the latest machines. I still go for these reasons, but now machines and software are improving in a more evolutionary fashion than the somewhat revolutionary changes at the beginning. Going from serial computers to vector or parallel ones was more exciting and groundbreaking than going from 100,000 cores to 1,000,000, though the latter is still challenging. Some things have stayed the same: the parties are still good, and companies are still entering and leaving the supercomputing area. For quite some time, if I brought home a coffee mug from a company, the company would go bankrupt in a few years. More recently, IBM developed the BlueGene series of machines, and grabbed the #1 spot in the top 500 rating of machines, but then dropped out of the market because it wasn’t selling enough machines to recoup the tremendous design cost.

One thing that has happened in computing field, not just the conference, is that scientific computing has a far smaller share of the market, even if you only consider the market for large systems. There have always been large database systems in corporations, but data analytics has greatly expanded the possibilities for profit, and hence there is more investment.

Question: What do you predict for the future of supercomputing?

Stout: The most “super” computers aren’t really single computers, but systems such as Google where they are continually processing a vast number of queries, answering them in fractions of a second by using sophisticated algorithms that combining myriad sources from throughout the world, all run on highly tuned systems that keep running even though they have so many components that they are always having to deal with faulty ones. The production users of supercomputers tend to submit a job, let it run for a long time, analyze the results (perhaps using sophisticated graphics), fix some errors or change some parameters, repeat. This isn’t the same as systems which are constantly ingesting data, analyzing it using algorithms that incorporate learning components, responding to increasingly complex queries. Academics, including some at U-M, are involved in this, but it is difficult to create even a scaled down version of a complete system in an academic computing center. You can view IBM’s Watson as being in this arena, and IBM is now betting that Watson will be a large part of its future.

Here’s an interesting cycle in computing: for over a decade some computational scientists and engineers have been using GPUs (graphics processing units). They are very difficult to use, and only applicable to certain types of problems, but inexpensive in terms of flops/$. However, many scientific computations require double precession arithmetic, which isn’t needed for graphics. Companies like NVIDIA, responding to the scientific computing market, began producing more expensive GPUs with double precision, and now systems such as U-M’s Flux computing cluster include GPUs on some of their boards.

However, there is a very rapidly growing demand for “deep learning.” The computationally intensive components of this can be run on GPUs relatively easily, but they don’t need double precision, just speed and plenty of parallelism. This summer NVIDIA released a new high-end chip with good double precision performance, but also added half precision, since that is all that is needed for deep learning. Deep learning might well surpass scientific computing as a GPU market.

 [NOTE: Visit the University of Michigan at SC16 at booth 1543.]

 

U-M prepares for SC16 conference in Salt Lake City

By | Uncategorized | No Comments

University of Michigan researchers and professional research IT staff will participate in the SC16 conference in Salt Lake City from Nov. 13-17 in a number of ways, including demonstrations, presentations and tutorials. Please join us at booth 1543 if you’re at the conference, or at one of the following events:

Sunday, Nov. 13
8:30 a.m. – 5 p.m.: Quentin Stout (EECS) and Christiane Jablonowski (CLASP) will teach the “Parallel Computing 101” tutorial.

Monday, Nov. 14 through Thursday, Nov. 17
U-M will exhibit at booth #1543 alongside Michigan State University. The booth will include an ongoing demonstration of the OSiRIS networking and storage project; information on the Yottabyte Research Cloud; and a presentation on ConFlux.

Tuesday, Nov. 15
10:30 a.m.: Todd Raeker, Research Technology Consultant, ARC-TS, will give a talk on ConFlux at the NVIDIA booth (#2231).
11 a.m.: Project PI Shawn McKee (Physics) will give a presentation on OSiRIS at the U-M booth (#1543).
2:15 p.m.: Nilmini Abeyratne, a Ph.D student in computer science, will present “Low Design-Risk Checkpointing Storage Solution for Exascale Supercomputers” at the Doctoral Showcase.
1 – 5 p.m.: Todd Raeker, Research Technology Consultant, ARC-TS, will participate in the IBM Power8 University Group Meeting.
3 p.m.: Representatives from Yottabyte and ARC-TS will give a presentation on the Yottabyte Research Cloud.
3:30 – 5 p.m., Sharon Broude Geva, Director of Advanced Research Computing, will participate in a panel titled “HPC Workforce Development: How Do We Find Them, Recruit Them, and Teach Them to Be Today’s Practitioners and Tomorrow’s Leaders?

Wednesday, Nov. 16
10 a.m.: Representatives from Yottabyte and ARC-TS will give a presentation on the Yottabyte Research Cloud.
11 a.m.: Ben Meekhof, HPC Storage Administrator, ARC-TS, will give a presentation on OSiRIS at the U-M booth (#1543).
1 p.m.: Todd Raeker, Research Technology Consultant, ARC-TS, will give a talk on ConFlux at the U-M booth (#1543).
5:15 – 7 p.m.: Ben Meekhof, HPC Storage Administrator, ARC-TS, will present at a “Birds of a Feather” meeting on “Ceph in HPC Environments.”

Thursday, Nov. 17
11 a.m.: Project PI Shawn McKee (Physics) will give a presentation on OSiRIS at the U-M booth (#1543).
1 p.m.: Todd Raeker, Research Technology Consultant, ARC-TS, will give a talk on ConFlux at the U-M booth (#1543).

Ann Arbor Deep Learning annual event — Nov. 12

By | Uncategorized | No Comments

a2-dlearn2016 is an annual event bringing together deep learning enthusiasts, researchers and practitioners from a variety of backgrounds.

MIDAS is proud to co-sponsor the event, which began last year as a collaboration between the Ann Arbor – Natural Language Processing and Machine Learning: Data, Science and Industry meetup groups.

The event will include speakers from the University of Michigan, University of Toronto, Toyota Research Institute and MDA Information Systems.

Please visit the event website for more information. Registration is required as space is limited.

Info Session: Data Science Services at U-M — Nov. 1

By | Uncategorized | No Comments

Representatives of Consulting for Statistics, Computing and Analytics Research (CSCAR) and the U-M Library (UML) will give an overview of services that are now available to support data-intensive research on campus.  As part of the U-M Data Science Initiative, CSCAR and UML are expanding their scopes and adding capacity to support a wide range of research involving data and computation.  This includes consulting, workshops, and training designed to meet basic and advanced needs in data management and analysis, as well as specialized support for areas such as remote sensing and geospatial analyses, and a funding program for dataset acquisitions.  Many of these services are available free of charge to U-M researchers.  

This event will begin with overview presentations about CSCAR and Library system data services.  There will also be opportunities for researchers to discuss individualized partnerships with CSCAR and UML to advance specific data-intensive projects.  Faculty, staff, and students are welcome to attend.  

Time/Date: 4-5 p.m., November 1,
Location: Earl Lewis Room, Rackham Building

Research highlights: A new era in disaster research

By | Uncategorized | No Comments

By Bob Brustman, U-M Civil and Environmental Engineering Department

University of Michigan researchers have received a $2.5 million NSF grant to develop a computational model that is hoped to significantly advance natural hazards engineering and disaster science.

Natural hazards engineers study earthquakes, tornadoes, hurricanes, tsunamis, landslides, and other disasters. They work to better understand the causes and effects of these phenomena on cities, homes, and infrastructure and develop strategies to save lives and mitigate damage.

Sherif El-Tawil

Sherif El-Tawil

Sherif El-Tawil, the lead PI for the project, is a structural engineer interested in how buildings behave, particularly in natural or man-made disasters. He’s developed 3D models and simulators that show precisely what happens in a building if a particular column or wall is destroyed during an extreme event.

On the project team are Jason McCormick, an earthquake engineering expert, Seymour Spence, who has expertise in wind engineering, and Benigno Aguirre, who is a social scientist interested in how people behave during catastrophes. The rest of the team includes. Vineet Kamat, Carol Menassa, and Atul Prakash, who will develop the simulation techniques used in the project.

The researchers of this newly funded project are creating a computational framework, using the Flux high performance computing cluster, that will define a set of standards for disaster researchers to use when constructing their models, enabling simulation models to work together.

El-Tawil explains: “Disaster research is a thriving area because disasters affect so many people worldwide and there is a lot we can do to reduce loss of life and damage to our civil infrastructure.”

“Lots of researchers study disasters, including engineers like me, but also social scientists, economists, doctors, and others. But all of the studies are essentially niche studies, belonging in the field of the researchers. Our objective is to develop computational standards so that social scientists, engineers, economists, doctors, first responders, and everyone else can produce simulators that interact together in a large, all-encompassing simulation of a disaster scenario. Think of it as the civilian equivalent of a war games simulator.”

el-tawil-nsf“Developing this common computational language will allow completely new studies to occur. Someone might look at the effects of an earthquake on a particular town and its citizens and then the subsequent effects of infectious diseases. With a common language, we can really examine the cascading and potentially out-of-control effects that occur during catastrophic events.”

Beyond developing the computational standards, they hope to create something like an app store through which researchers can share their simulation models and foster new collaborations and new areas of research. 

The grant also includes funding for a programmer housed at Advanced Research Computing (ARC) that will become a shared resource for the rest of campus. The Michigan Institute for Computational Discovery and Engineering (MICDE) provided support for the grant submission, and will continue to do so post-award.

The project brings together an experienced team with expertise in engineering, social science, and computer science. Six of the seven core members are from the University of Michigan and the seventh is from the University of Delaware.

Team members:

  • Benigno Aguirre, professor, Disaster Research Center, University of Delaware
  • Sherif El-Tawil, professor, Department of Civil and Environmental Engineering, University of Michigan
  • Vineet Kamat, professor, Department of Civil and Environmental Engineering, University of Michigan
  • Jason McCormick, associate professor, Department of Civil and Environmental Engineering, University of Michigan
  • Carol Menassa, associate professor, Department of Civil and Environmental Engineering, University of Michigan
  • Atul Prakash, professor, Department of Electrical Engineering and Computer Science, University of Michigan
  • Seymour Spence, assistant professor, Department of Civil and Environmental Engineering, University of Michigan