University of Michigan’s Ph.D. in Scientific Computing: A history of supporting research through education

By | Educational, Feature

#Computationalscience everywhere!

Left side, 2167 configuration console for the IBM/System 360 Model 67-2 (duplex) at the University of Michigan, c. 1969 [Picture by Scott Gerstenberger – Scott Gerstenberger, Public Domain]

The University of Michigan’s joint Ph.D. program in Scientific Computing recently achieved a record enrollment of 137 students. Between 2015, when 15 students were enrolled -mainly from the Colleges of Engineering and Literature, Science and the Arts- and today, the program has witnessed an explosive growth of interest on the part of U-M students. The program now has students enrolled from over 30 departments spanning 8 different schools and colleges, and more than 130 students have graduated in the last 31 years, including 17 students to-date in 2020.

This popularity is emblematic of the dominant role that computation plays in the world today. With the breakneck pace at which new hardware and software architectures are being developed, the boom in simulation-based research in a growing number of disciplines, and the initiatives in data and computational sciences implemented at U-M in the last few years, including the establishment of the Michigan Institute for Computational Discovery & Engineering, and the Michigan Institute for Data Science (MIDAS), it may seem only natural that scientific computing should attract this level of interest. However, like all exceptionally successful undertakings, it owes a great deal to its past. We reached back more than three decades to piece together the history of the Ph.D. in Scientific Computing at U-M.

The broader history of computational science and high performance computing at the University of Michigan is rich and extensive. U-M has been at the forefront of Cyberinfrastructure research for many decades, marked by the acquisition of U-M’s first virtual memory computer in 1967, an IBM 360/67, one of the first computers of its kind in the world. This milestone was followed by many others, including further hardware acquisitions and establishment of new units to support advanced research computing. An important early step was taken in 1985 when the College of Engineering established the Laboratory for Scientific Computation (LaSC). LaSC’s goal was to foster and promote the use of scientific computation in research and instruction at U-M. During those years, several reports from national study committees established computational science as the third pillar of scientific methodology, along with theory and experimentation. Faculty members of LaSC, who were at the forefront of driving these trends  recognized that any initiative in this field needed to include a robust student training program. 

left: Prof. Kenneth Powell (Aerospace Engineering), director of the Ph.D. in Scientific Computing program since 2005; right: Prof. William Martin (Nuclear Eng. and Rad. Sciences), director of the program from 1989 to 2004.

Prominent at that time in LaSC were Prof. William “Bill” Martin (Nuclear Engineering and Radiological Sciences – NERS), the laboratory’s director, Prof. John Boyd (Atmospheric, Oceanic and Space Sciences), the laboratory’s associate director, and Prof. Edward Larsen (NERS), who was hired as part of the College of Engineering’s initiative to move aggressively in the area of scientific computing. Together, they designed a graduate academic program with the goal of giving students a more comprehensive training in numerical analysis and computer science than is typically possible within standard disciplinary programs housed within individual departments and schools. The fundamental idea was that, to excel in computational science and engineering, one must have a thorough understanding of the mathematical and physical problems to be solved, expertise in  the methodologies and algorithms, and a foundation in computer science to be able to apply this arsenal of techniques on modern computer platforms. The need for a thorough understanding of the physical problems led directly to the requirement that students had to be enrolled in a traditional Rackham degree program (i.e., a home department), while the need for mathematical underpinning and knowledge of algorithms and computer science topics led to the requirements for courses in numerical analysis, parallel algorithms, and related topics. The PhD in Scientific Computing program was approved by the State of Michigan in 1988, and enrolled its first students in 1989. This was well in advance of a wider recognition of the centrality of computation in academia and industry. It is true today, as it was in 1988, that students can apply to the PhD in Scientific Computing program from any Rackham-recognized PhD program at the UM. This unique and flexible administrative structure has enabled the rapid growth experienced in recent years as scientific computing has become an indispensable tool in many fields of academic endeavor. 

Prof. Quentin Stout, director of the Center for Parallel Computing 1992-2001 [Picture source: NASA Insights 1998]

The oversight of the degree program has evolved over the years as administrative structures around scientific computing have shifted. Regardless of its administrative home, the program has always been organized under the Rackham School of Graduate Studies. Originally, the College of Engineering had oversight of the program, with Prof. Martin appointed as director, and with guidance from the LaSC Education Committee. This setup continued through the merger of LaSC and the Center for Parallel Computing1 into the Center for Advanced Computing in 2001. In 2005, Prof. Kenneth Powell (Aerospace Engineering) was named director of the program succeeding Prof. Martin, and has continued in the role since. In 2008, the Office of Research Cyberinfrastructure (ORCI) was established, and the oversight of the program changed to the U-M Office of Research. In 2013, when ORCI was re-named as Advanced Research Computing, and the Michigan Institute for Computational Discovery & Engineering (MICDE) was born, oversight was transferred to MICDE.

Since its inception, the program has been described as intended for students who will make intensive use of large-scale computation, computational methods or algorithms in their doctoral studies. Although the requirements and goals of the program have not  changed in 31 years, the research applications, the algorithms and methodologies, and the computer platforms have been in constant evolution. Naturally, the courses offered in support of the program have followed closely. In 1989 the core research areas behind the program were computational fluid dynamics, advanced computer architectures, and particle transport, with the majority of the students coming from engineering, and mathematics. Still, students working in areas where computation was less recognized, such as AIDS transmission or social research projects, also were enrolled. Over the next two decades, the tremendous increase in simulation-based research by the faculty in engineering and physical sciences added many other focus areas, including material science, astronomy, and high energy physics, to name just a few. This growth added a new driver as data-intensive research gained importance in those fields. 

Prof. Suzanne Weekes, Associate Dean of Undergraduate Studies, ad interim, and Professor of Mathematical Sciences at Worcester Polytechnic Institute (U-M 1995, Mathematics and Scientific Computing) [Picture source: SIAM News Sept. 2020]

Several faculty members have had an important role shaping the program, by offering fundamental courses and providing mentorship. Notably, Prof. Quentin Stout, from Computer Science and Engineering, has had a prominent role in the program. He was the founding director of the Center for Parallel Computing, which  provided the basis for subsequent units in this sphere at U-M. He also developed, and has been teaching, Parallel Computing since 1985, innovating its curriculum to remain at the cutting edge of the current techniques, important aspects of which have been based on his own research. Other foundational courses, such as the Department of Mathematics’ Numerical Methods for Scientific Computing I & II and Numerical Linear Algebra have been offered for more than 30 years. More recently the Department of Physics course, Computational Physics, and the College of Engineering course, Methods and Practice of Scientific Computing, along with an array of courses in machine learning, have played prominent roles in transforming the curriculum in scientific computing as research in these areas has likewise redefined the field.

Unsurprisingly, the Ph.D. in Scientific Computing has produced many exceptional alumni. The first student graduated from the program in 1992, and notably for its time, two of the first four graduates were women, when gender imbalance was barely recognized. A majority of the program graduates went on to  positions in academia or the National Laboratories, with the rest working in varied fields in industry or government. Some of these outstanding alumni include Suzanne Weekes, U-M 1995 (Mathematics and Scientific Computing), currently Associate Dean of Undergraduate Studies, ad interim, and Professor of Mathematical Sciences at Worcester Polytechnic Institute. Prof. Weekes has recently been named SIAM executive director, and will start her new role on January 1, 2021.  Another alumna, Rona Oran, U-M 2014 (Space Science and Scientific Computing), is a computational plasma physicist at MIT and a member of the NASA team that is designing and planning a mission to the metal asteroid Psyche to be launched in 2020.

The current goal of the program is still founded on the original idea of strengthening the students’ foundations in methodology and computer science. The leadership of the program strives to bring computational science to more research fields, but importantly, aims to do so by enhancing diversity in the field. An important marker of U-M’s success on this front came in  2018 in the form of the Henry Luce Foundation’s award to the University of two Claire Boothe Luce Ph.D. fellowships for women to enroll in the Ph.D. in Scientific Computing. The program is committed to pursuing other such opportunities and creating an environment where students of all backgrounds and identities feel welcome and thrive.

1 In 1992 U-M was awarded a major equipment grant by the National Science Foundation to create a testbed of parallel computing architectures. The Center for Parallel Computing was established to operate the facility. The center installed and operated several different parallel computers over the years, including KSR-1, KSR-2, Convex Exemplar, SGI PowerChallenge, IBM SP2, and AMD and Apple clusters.

Reducing lung cancer mortality through modeling and simulations

By | Feature, Research

Lung cancer remains the leading cause of cancer related mortality in the US, and globally, accounting for 1.8 million deaths annually. Many of these deaths are preventable by the implementation of prevention strategies, including tobacco control policies and lung cancer screening recommendations, and by improvements in lung cancer treatment.  In the US, these policies have generally been implemented based on the analysis and outcomes of the population as a whole, although data analyses have shown that smoking and lung cancer rates, and access to healthcare and interventions, vary significantly by education, income, and race/ethnicity.

The Cancer Intervention and Surveillance Modeling Network (CISNET) Lung Working Group (LWG), led by Rafael Meza, associate professor of Epidemiology from the School of Public Health and MICDE member, has been awarded a new $8.5M grant to investigate the synergistic impacts of tobacco control policies, lung cancer screening and treatment interventions in the US and in middle-income nations. For the past 15 years, the CISNET LWG has contributed to the development of US national strategies for reducing the lung cancer burden by quantifying, through modeling and simulation, the impact of tobacco control on smoking, lung cancer, and overall mortality, as well as the population benefits and harms of lung cancer screening. This new grant will allow the group to expand their work to consider the impact of treatment improvements, including targeted therapies and immunotherapies,  and the synergies between treatment and prevention interventions. It also will enable the researchers to continue their work in addressing smoking and lung cancer disparities. The consortium uses a comparative modeling approach, where multiple, but distinct, models use the same data inputs, and aim to answer a common question with different approaches. This allows the group to assess the strengths and weaknesses of the different models, and aid the decision making process.

Established in 2000, CISNET is a consortium of NCI-sponsored investigators who use modeling and simulation to improve their understanding of cancer control interventions in prevention, screening, and treatment and their effects on population trends in incidence and mortality. CISNET is committed to bringing the most sophisticated evidence-based planning tools to population health and public policy. These models have been used to guide public health research and priorities, and have aided the development of optimal cancer control strategies. Besides lung cancer, CISNET also includes breast, colon, cervical esophageal and prostate cancer groups. 

Modeling the transmission of infectious aerosols

By | Feature, Research

Inhalation of micron-sized droplets represents the dominant transmission mechanism for influenza and rhinovirus, and recent research shows that it is likely also the case for the novel coronavirus.  Increasing evidence suggests that the transmission of infectious aerosols is more complex than previously thought. Coughing, sneezing and even talking yield a gaseous flow field near the infected person that is dynamic and turbulent in nature. Existing models commonly employed in simulations of aerosol transmission attempt to represent the effect of turbulence using random walk models that are often phenomenological in nature, employing adjustable parameters and inherently assuming the turbulent fluctuations ‘felt’ by a droplet do not depend upon direction. To design physics-informed guidelines to minimize the spread of this virus, improved predictive modeling capabilities for effectively tracking the aerosol paths are needed. Dr. Aaron M. Lattanzi and Prof. Jesse Capecelatro, from Mechanical Engineering and MICDE are tackling this problem by focusing on mathematical modeling of aerosol dispersion. They derived analytical solutions for the mean-squared-displacement resulting from systems of stochastic differential equations. A key element of their methodology is that the solution connects stochastic theory inputs to statistics present in high-fidelity simulations or experiments, providing a framework for developing improved models.

Simple simulation of aerosol dispersion from a single-point source. The grey, cone-like surface is the approximation using Force Langevin (FL) theory and the colored particles are from integration of Newton’s equations with stochastic drag forces.

Prof. Capecelatro’s research group develops physics-based models and numerical algorithms to leverage supercomputers for prediction and optimization of the complex flows relevant to energy and the environment. The main focus of their research involves developing robust and scalable numerical tools to investigate the multiphysics and multiscale phenomena under various flow conditions, like those that they study here. They recently submitted their findings to the Journal of Fluid Mechanics, and are continuing to work on this problem hoping it will help understand the transmission of COVID-19 and therefore help optimize current guidelines.

Applications for MICDE Graduate Fellowships are being accepted

By | Feature

Since 2014, MICDE has offered top-off fellowships to current and prospective students whose research project involves the use and advancement of scientific computing techniques and practices. These fellowships, which carry a $4,000 stipend, are meant to augment other sources of funding. Applications from students enrolled in one of MICDE’s educational programs are being accepted.

The deadline to submit an application is Fri., June 19, 2020. See the Call for Applications.

group photo of 2019-20 fellows

2019-2020 MICDE Fellows (from left to right) Guodong Chen (Aero), Suyash Tandon (ME), Jiale Tan (Epidemiology), Chongxing Fan (ClaSp), Kelly Broen (Epidemiology), Bradley Dice (Physics), Liz Livingston (ME), Will Weaver (EEB), Yuan Yao (ME), Samuel Baltz (Pol Sci), Joe Hollowed (Physics), Minki Kim (ME), Allison Roessler (Chem), Fuming Chang (ClaSp), Maral Budak (Biophysics), Saibal De (Math), Xian Yu (IOE), Jiaming Zhang (Physics). [Not pictured: Thomas Waltmann (Physics), Anil Yildirim (Aero), and Jessica Conrad (IAM)]

U-M draws global attention for MOOC: Problem Solving using Computational Thinking

By | Educational, Feature, Research

Problem Solving using Computational Thinking, a Massive Open Online Course (MOOC) launched by the University of Michigan in November of 2019, has already drawn more than 1,200 learners from around the globe. The Michigan Institute for Computational Discovery & Engineering (MICDE) and the University of Michigan Center for Academic Innovation partnered to create this course. The idea for this MOOC arose from the team’s recognition of the ubiquity of computation. However, the developers were equally keen to distinguish this offering from MOOCs on programming and to instead highlight how broader computational thinking also makes its presence felt in somewhat unexpected domains.

Using computational thinking, the MOOC challenges learners with a series of real-world examples, including how it is possible to help plan and prepare for a flu season–a subject that has gained particular relevance in the months following the launch of this MOOC, track human rights violations or monitor the safety of crowds.

While enrollment numbers are encouraging, the work being done by learners within the MOOC is most inspiring. For their final project, learners have applied the computational thinking strategies discussed throughout the MOOC to a wide array of noble social problems in hopes of finding cogent solutions.

Not surprisingly, there have been several projects that seek to address challenges related to COVID 19.

The MOOC’s Epidemiology Case Study walks the student through the process of building a communicable disease transmission model.

One learner wrote: “For the final project, I am assuming the role of a member of the team responsible to combat COVID-19 from India and I have to decide on what should be our strategy to fight coronavirus in India, be it the extension of a lockdown or any other important decision related to this pandemic.”

In another project, a learner assuming the role of a Wuhan pathologist wrote that they must “decide what the Chinese government’s strategy against coronaviruses” should be.

Learners addressing today’s most pressing societal concerns, such as COVID-19, exemplifies the transformative potential of open-access, digital, and distance education made possible by a MOOC.

Across the board, the MOOC has received tremendously positive reviews, with an overall course rating of 5 out of 5 stars. One learner, in particular, wrote in their course review: “I really enjoyed this course! It got me prepared to study for an entry into a career working with computers!!” Another learner simply stated: “Fantastic, loved it!”

The developers of this MOOC are drawn from the School of Public Health, the College of Engineering, the School of Education and MICDE. Problem Solving using Computational Thinking is available in Coursera through Michigan Online. To learn more please visit online.umich.edu/courses/problem-solving-using-computational-thinking/.

U-M modeling epidemiologists helping navigate the COVID-19 pandemic

By | Feature, News, Research

[top] Screenshoot of the Michigan COVID-19 Modeling Dashboard (epimath.github.io/covid-19-modeling/); [bottom left] Marisa Eisenberg (Epidemiology, Complex Systems and Mathematics); [bottom right] Jonathan Zelner (Epidemiology).

The COVID-19 pandemic is producing massive amounts of information that more often than not lead to different interpretations. The accurate analysis of this daily input of data is crucial to predict possible outcomes and design solutions rapidly. These can only be achieved with expertise in modeling infectious diseases, and with the power of computational science theory and infrastructure. U-M’s Epidemiology Department, in the School of Public Health, has a very strong cohort of researchers who work on mathematically modeling the dynamics of infectious diseases, the analysis of these models, and large scale computer simulations — all to understand the spread and mitigation of pandemics. They are applying their long experience and expertise to the current COVID-19 outbreak, aiding the government make informed decisions, and helping media outlets produce accurate reports for the general public. Marisa Eisenberg, Associate Professor of Epidemiology, of Complex Systems, and of Mathematics, and her colleagues are using a differential equation transmission modeling approach to analyze scenarios and generate short-term forecasts for the COVID-19 epidemic in State of Michigan. They are communicating directly with the Michigan Department of Health and Human Services and providing critical tools, like the Michigan COVID-19 Modeling Dashboard, to inform decision-making. Prof. Eisenberg’s team is helping to forecast the numbers of laboratory-confirmed cases, fatalities, hospitalized patients, and hospital capacity issues (such as ICU beds needed), and examining how social distancing can impact the spread of the epidemic. Prof. Jonathan Zelner, whose research is focused on using spatial and social network analysis and dynamic modeling to prevent infectious diseases, is part of a group helping map the outbreak in Michigan. He also has provided valuable insights to journalists contributing to a better understanding of the situation, including what made New York City so vulnerable to the coronavirus (NYT), the role of wealth inequality during epidemics (CNBC) and what professions and communities are particularly vulnerable to infection (NG). 

Professors Eisenberg and Zelner are not alone in this fight. Many more researchers from U-M’s School of Public Health and throughout campus have risen to the challenges posed by this pandemic. 

Combat COVID-19 using newly available HPC resources: COVID-19 High Performance Computing Consortium

By | Feature, HPC, News, Research

COVID-19 High Performance Computing Consortium

On March 23, 3030 the White House announced the launch of a new partnership that aims to unleash U.S. supercomputing resources to fight COVID-19: the COVID-19 High Performance Computing Consortium. The goal of the Consortium is to bring together the Federal government, industry, and academic leaders to provide access to the world’s most powerful high-performance computing resources in support of COVID-19 research. The access to these resources has the potential to significantly advance the pace of scientific discovery in the fight to stop the virus.

To request access to resources of the COVID-19 HPC Consortium, you must prepare a description, no longer than two pages, of your proposed work. To ensure your request is directed to the appropriate resource(s), your description should include the following sections. Do not include any proprietary information in proposals, since your request will be reviewed by staff from a number of consortium sites. It is expected that teams who receive Consortium access will publish their results in the open scientific literature.

Learn more at https://covid19-hpc.mybluemix.net .

 

 

 

 

Learn more about the COVID-19 outbreak through a panel of experts from the Society of Risk Analysis

By | Feature, Happenings, News

Seth Guikema, Professor of Industrial & Operations Engineering, MICDE affiliated faculty, and President of the Society of Risk Analysis moderated the webinar on Coronavirus: Risk Analysis Perspectives on COVID-19 Outbreak on Thursday, March 12, 2020. The webinar featured a panel of risk experts from the Society of Risk Analysis. If you missed the webinar yesterday you can still watch a recording of the panel discussion online.

 

2020 Argonne Training Program on Extreme-Scale Computing (ATPESC)

By | Feature, SC2 jobs

2020 Argonne Training Program on Extreme-Scale Computing (ATPESC)

Application deadline: March 2, 2020.

There are no fees to participate in ATPESC. Domestic airfare, meals, and lodging are also provided.

Apply for an opportunity to learn the tools and techniques needed to carry out scientific computing research on the world’s most powerful supercomputers. ATPESC participants will be granted access to DOE’s leadership-class systems at the ALCF, OLCF, and NERSC. This year’s program will take place July 26–August 7, 2020.

PROGRAM CURRICULUM

Renowned scientists and leading HPC experts will serve as lecturers and guide the hands-on sessions. The core curriculum will cover:

  • Hardware architectures
  • Programming models and languages
  • Data-intensive computing and I/O
  • Visualization and data analysis
  • Numerical algorithms and software for extreme-scale science
  • Performance tools and debuggers
  • Software productivity
  • Machine learning and deep learning for science

ELIGIBILITY AND APPLICATION

Doctoral students, postdocs, and computational scientists interested in attending ATPESC can review eligibility and application details on the website.

New MOOC in Computational Thinking has launched!

By | Educational, Feature, Happenings

The Michigan Institute for Computational Discovery & Engineering and the University of Michigan Center for Academic Innovation have partnered to launch a Massive Open Online Course (MOOC) titled Problem Solving using Computational Thinking. The idea for this MOOC arose from the team’s recognition of the ubiquity of computation. However, the developers were equally keen to distinguish this offering from MOOCs on programming, and to instead highlight how broader computational thinking also makes its presence felt in somewhat unexpected domains. The MOOC is organized in a series of real-world examples that includes how, using computational thinking, it is possible to help plan and prepare for a flu season, track human rights violations or monitor the safety of crowds. The process of computational thinking that this MOOC focuses on ranges from problem identification, through abstraction to evaluating solutions. Problem Solving using Computational Thinking seeks to introduce students and teachers to the systematic thinking needed to conceptualize a problem with the intent of eventually using some computational tools to solve it.

The developers of this MOOC are drawn from the School of Public Health, the College of Engineering, the School of Education and MICDE. Problem Solving using Computational Thinking is available in Coursera through Michigan Online. To learn more please visit online.umich.edu/courses/problem-solving-using-computational-thinking/.