Computationally intensive Research in the MGHPCC Consortium
High-performance computing (HPC) systems that bring together large numbers of CPUs and massive amounts of storage, are playing a transformative role in research. For example, high performance computers are used to create, simulate and solve models of complex phenomena that range from protein structure, to turbulent fluid flows, to the dynamics of the earth’s atmosphere, to pathways of human social interaction, to galactic evolution. These models provide new perspectives and new ways of making discoveries. The following is a small sample of the computationally-intensive research being conducted at the MGHPCC universities. Clicking on our RESEARCH tag gives a more detailed view of current activity.
Stronger. Faster. Better.
Pushing performance through computer architecture and algorithm development: Northeastern students present their innovative research at HPC Day 2018. MORE
Sensing Subduction Zones
UMass geoscientist Haiyong Gao uses MGHPCC to handle the huge sets of sensor data needed for the advanced methods of seismic imaging she has developed to help her understand the movement of Earth’s tectonic plates. MORE
Data Visualization using Climate Reanalyzer
A team from the University of Maine uses a Northeast Cyberteam Program seed grant to upgrade a public climate data vizualisation tool developed at the U Maine Climate Change Institute. MORE
Getting to Grips with Glassy Materials
Jon Machta, an emeritus Professor of Physics at UMass Amherst, works in the area of theoretical condensed matter and statistical physics. His group has been using MGHPCC to explore the physics of glasses, materials that are neither liquid nor solid but share properties of both. MORE
Modeling Molecular Engines
Paul Whitford, an Assistant Professor of Physics at Northeastern University, uses high performance computing and the facilities at the MGHPCC to study the dynamics of biological systems to understand the physical principles that govern the dynamics of cells. MORE
Forest Mapping: When the budworms come to dinner
Motivated by an eastern spruce budworm outbreak traveling down from Canada, researchers in the School of Forest Resources at the University of Maine and colleagues in the U Maine Advanced Computing Group, catalyzed by a seed grant from the Northeast Cyberteam Program, have been applying machine learning techniques to map the evolving state of the forest to provide accurate and up-to-date information on forest as the outbreak develops. MORE
Exploring Thermoelectric Behavior at the Nanoscale
Zlatan Aksamija, an Assistant Professor of Electrical and Computer Engineering at the University of Massachusetts Amherst, uses computers at the MGHPCC to carry out nanomolecular materials modeling experiments exploring the thermoelectric behavior of materials for use in energy applications.MORE
The Trickiness of Talking to Computers
James Glass is a senior research scientist at the Massachusetts Institute of Technology. Glass leads the Spoken Language Systems Group in the Computer Science and Artificial Intelligence Laboratory (CSAIL.) His research is focused on automatic speech recognition, unsupervised speech processing, and spoken language understanding. This past spring, assisted by graduate student David Harwath, Glass was the instructor for MIT’s 6.345/HST.728 Automatic Speech Recognition class but this year, for the first time, students had the option of using high performance computing resources at the MGHPCC to facilitate their work. MORE
A Genomic Take of Geobiology
Abigail Caron, a postgraduate researcher in Prof Greg Fournier’s Geobiology Lab at the Massachusetts Institute of Technology (MIT) uses a computer cluster housed at the MGHPCC to run genetic analyses on different bacteria looking for instances of horizontal gene transfer, and mapping these events across many lineages in work that seeks to calibrate the ancient history of life on Earth. MORE
Lessons in a Virtual Test Tube
Dr. Saritha Nellutla, a magnetic materials scientist at Bridgewater State University, uses C3DDB at the MGHPCC to carry out virtual lab experiments in the chemistry course she teaches. MORE
Neural Networks & Earthquakes
Brendan Meade’s Group at Harvard, has been using the MGHPCC to speed-up earthquake cycle modeling using neural networks. The potential for loss of property and life, has made earthquake forecasting and prediction an active area of research for statisticians and earth scientists. While it is not currently possible to make deterministic predictions of when and where earthquakes will happen, the Meade Group’s new technique bring closer the day when the behavior of a seismically active region can me modeled quickly enough to be able to provide meaningful lifesaving guidance to at risk populations. MORE
Small Stars, Smaller Planets, Big Computing
Mark Veyette, a PhD student at Boston University, has been using the MGHPCC to run state-of-the-art models of M dwarf star atmospheres to better understand their composition characteristics and how that relates to the types of planets that form around them. MORE
Award Winning Innovation in Analytics
The MGHPCC is host to the IOMICS Corps. FUSION Analytics PlatformTM, a cloud-based software system for prescriptive analytics and rapid prototyping of advanced decision models for use in chemical engineering, medical research, and clinical care. MORE
Teaching Computers to Identify Odors
Harvard professor of molecular and cellular biology Venkatesh Murthy is using computers at the MGHPCC and machine learning algorithms to “train” the computers to recognize neural patterns associated with different scents. MORE
Grass to Gas
University of Massachusetts Amherst computational chemist Scott Auerbach is using the MGHPCC in research helping him understand and optimize the process of producing fuels such as gasoline from plant biomass instead of from petroleum. MORE
From Games to Brains
Researchers from Northeastern University are using computers at the MGHPCC to improve the performance of a popular medical imaging tool which estimates 3D light distribution in biological tissue using GPU technology to simultaneously simulate the paths of large numbers of independent photons. MORE
The Trouble with Turbulence
Seyedeh Mahnaz Modirkhazeni, a PhD candidate attached to the Re-Engineering Energy Laboratory in the Mechanical Engineering Department at UMass Lowell, has been using computers at the MGHPCC to explore a new technique for modeling non-equilibrium turbulent flows of the kind generated in direct current (DC) arc plasma torches; a tool found at the heart of many industrial thermal plasma processes among them metal cutting and welding, thermal plasma chemical vapor deposition (CVD), metal melting and re-melting, waste treatment, and gas production. MORE
A New Twist
Polymer scientist Greg Grayson at Umass Amherst has been using MGHPCC to understand what sets the final size and shape of chiral filament bundles (many stranded, self-twisting, yarn like structures) implicated in such diseases as Alzheimer’s, Parkinson’s, and Sickle Cell Anemia. MORE
Heading Off Head Blight
Head blight caused by Fusarium graminearum threatens worldwide wheat production, resulting in both yield loss and mycotoxin contamination. Molecular biologists at UMass Amherst are using MGHPCC to understand pathogenicity at the systems level with the goal of developing novel disease control strategies. MORE
A Little Bit of This…
A Little Bit of That…
The MGHPCC helps Umass Chemist Dr. Dhandapani Venkataraman as he develops a new process for fabricating binary nanoparticle glass with myriad applications in solar cells, batteries, sensors, and beyond. Says Venkataraman: “A new paradigm for hierarchical molecular self-assembly and materials design in which the chemist not only assembles the molecules but then goes on to assemble the molecular assemblies!” MORE
Up in the Air
Chien Wang is a senior research scientist in the Department of Earth, Atmospheric and Planetary Sciences at MIT associated with MIT’s Center for Global Change Science and the Joint Program in the Science and Policy of Global Change. Wang and his group develop and use complex computer models to explore how atmospheric aerosols impact climate. MORE
Looking Like an Alien!
Paul Alba, an astronomy student at Boston University, has been using MGHPCC to answer the question: “What would a distant alien civilisation learn if they tried to study Saturn with the same methods we currently apply to exoplanets?” MORE
FloDesign Sonics is using MGHPCC to understand micron sized particles’ movement inside 3D acoustic fields. Such types of calculations involve solving trajectories of tens of thousands of particles and their effect on the flow field along with particle-particle collisions. Due to the complex nature of such computations, a supercomputer like MHGPCC is an ideal resource. The resulting data helps them understand the physics behind acoustical trapping, clustering and separation of particles from the bulk fluid and the type of acoustic field and geometry required for a more efficient process. This helps in optimizing their systems and aids product development. MORE
Designing Secure Computer Systems
Cyber security and the design of systems to protect data from attack by malicious entities, presents an enormous challenge. Computer architect Srini Devedas, a professor in the Computer Science and Artificial Intelligence Laboratory (CSAIL) at MIT, collaborates with software researchers at MIT and other universities, to meld secure architecture with secure compiler and operating system software, to protect data from the ground up MORE
Modern Neuroscience demands Big Data
Scientists at Harvard University are using the MGHPCC for their research to uncover the mysteries of the brain. Whether studying the connection and wiring of individual neurons at the micro-scale to exploring the interaction of brain areas across thousands of people using magnetic resonance imaging technologies at the macro-scale, either way, progress is predicated on efficient handling of big data. MORE
Modeling electron excitation in organic photovoltaic materials
Theoretical chemist Adam Willard, Assistant Professor at the Massachusetts Institute of Technology, and his group are using the MGHPCC facility in their computer modeling work, which seeks to understand the molecular-scale behavior of photonically excited-electrons in organic photovoltaic materials, which can be used to build thin-film solar panels made out of semi-conducting plastic MORE
ATLAS Northeast Tier 2 center.
ATLAS is a particle physics experiment at the Large Hadron Collider at CERN that is searching for new discoveries in the head-on collisions of protons of extraordinarily high energy. ATLAS will learn about the basic forces that have shaped our Universe since the beginning of time and that will determine its fate. ATLAS was one of the two experiments that recently observed the Higgs boson. The MGHPCC houses the computing facilities used by researchers at BU and Harvard involved with the project MORE
Toward a Smarter Greener Grid
Building a Smarter, Greener Grid. UMass Amherst computer scientist Prashant Shenoy and electrical and computer engineer David Irwin are leading a team of researchers focused on analyzing smart meters and other tools that could transform the way energy is utilized, monitored, and controlled in America’s buildings, which currently comprise 75% of the national electric bill. Along with partner Holyoke Gas and Electric (HG&E), Shenoy and his team are analyzing data from 18,000 smart meters in the city of Holyoke. The analysis will be used by projects that include design of smarter thermostats, a web-based solar predictor, and an electrical reserve battery. MORE
Strength and Fracture Mechanisms of Hierarchical Biological Materials.
Alain Karma (Northeastern), and Markus Buehler (MIT) are using high performance computers in a multiscale modeling effort that incorporates length-scales from the atom to the structural scale to explore and test how biological materials like bone or nacre can, despite their apparent fragility, resist breakage. The multiscale computational approach developed in this project aims to permit an exploration of fundamental mechanisms that control the strength of biological materials, and at the same time, create a platform for the biomimetic (literally nature mimicking) design of nanostructured composite engineering materials for structural and medical applications. MORE
Science and Policy of Global Change.
Understanding the complex, long-term changes in our land, air and water requires breakthroughs in measurement, modeling and prediction. Responding to these changes requires innovative policies that comprehend agriculture, energy needs, trade and finance — along with the political and communications savvy to organize a genuinely global approach. The Joint Program on the Science and Policy of Global Change is MIT’s response to these research, analysis, and public education challenges. At the heart of much of the Program’s work lies MIT’s Integrated Global System Model (IGSM), a linked set of computer models designed to simulate the global environmental changes that arise as a result of human causes. In this way, it explores the interplay between Earth systems and human activity. MORE
Multi-Scale Ocean Modeling
An MGHPCC seed fund award allowed Professor Pierre Lermusiaux and collaborators to develop multi-scale models of the marine environment off the New England coast designed to talk to one-another and interface with real-time observations MORE
Designing Cloud and Big Data Platforms for Scientific and HPC Applications.
Orran Krieger (BU), Peter Desnoyers (Northeastern), Manuel Garber (UMass Worcester), and Prashant Shenoy (UMass Amherst) are exploring ways to better tailor today’s cloud and big data platforms for Big Data – the massive volume of information so many disciplines from the life-sciences, through the physical-sciences, to the social sciences now generate. With an eye to genomics and sequencing applications in the life-sciences, real-time data mining of smart grids, even weather nowcasting (using weather data to develop short-term personalized forecasts), the team is looking at ways to re-architect big data processing methods and cloud platforms to better meet the needs of data intensive scientific applications. MORE
Informing Economic Policy Through Computation.
The Chetty research group, which is part of the Lab for Economic Applications and Policy (LEAP) at Harvard, focuses on understanding how to improve government policy to spark economic growth and reduce poverty. The group analyzes the effects of government policies using mathematical models of the economy and tests these models using large datasets on households and firms. Current questions include the long term economic effects of tax policies and government spending, the effect of increased transparency and salience on anti-poverty programs, and the economic effects of changing social norms. These theoretical and empirical research projects use high performance computers operated by the Harvard Research Computing Group to run complex dynamic models and analyze datasets that contain billions of observations. MORE
Robert Fisher, a professor of astrophysics at UMass Dartmouth, uses high performance computing to model and make predications about the behavior of supernovae. Supernovae are the incredible explosions that take place at the end of a star’s life cycle, and are among the brightest objects in the sky. They have been an object of fascination for astronomers and astrophysicists for many centuries. Fisher models their behavior using large computer simulations which simultaneously solve the complex physics equations that make up a detailed theoretical description of a star or star pair. MORE
Exploring the Dark Matter Substructure of Milky Way Galaxies.
Anna Frebel (MIT), Edmund Bertschinger (MIT), and Lars Hernquist (Harvard) are using astronomical observations and the fundamental laws of physics to improve our understanding of how galaxies assemble, their origins, and what controls their characteristics. With the advent of computers, astrophysicists have been able to build increasingly sophisticated models of the growth of agglomerations such as the Milky Way through so-called N-body simulations. In this project, the investigators seek to leverage High Performance Computing to allow them to run not one but multiple N-body problems in parallel, enabling them to apply statistical approaches to uncovering the underlying physical processes at work. It is hoped this will lead to the conclusive confirmation of the details of the assembly history of the Milky Way as well as a better understanding of general galaxy behavior.
Automated real-time medical imaging analysis.
Patricia Ellen Grant (Harvard Medical School & Children’s Hospital Boston), and Jonathan Appavoo (Boston University) are using high-performance computing to make real-time radiological image analysis easier and less costly to use. MORE
Organic Light Harvesting Antennas.
Typically plants need an average of at least a few hundred watts per square meter of solar radiation to survive. At depths of a few hundred meters down in the ocean, the amount of solar radiation decays to around a millionth of the surface intensity so that photosynthesizing organisms at such depths only receive 10-4 Wm-2 of radiation. MGHPCC seed fund awardees Alán Aspuru-Guzik (Quantum Chemistry, Harvard) and Alfredo Alexander-Katz (Materials Science, MIT) are studying the properties of a photosynthetic bacteria (Chlorobium Tepidum, also known as Green sulphur bacteria) that manages to survive in this low energy environment. Their work seeks to understand the interplay between biomolecular structure and efficient light harvesting on the quantum scale and how this could lead to radical new approaches to energy generation. MORE
Computational Identification of Outcome-Associated DNA Alterations in Neuroblastoma
Stefano Monti (BU) and Roberto Chiarlie (Harvard) are investigating ways to improve the effectiveness of treatments for Neuroblastoma, one of the most common tumor-causing cancers in children. Clinically, neuroblastoma is classified in 4 stages, with increasingly poor prognosis. However a significant subset of stage 4 patients (the most advanced stage) can be cured and may even regress spontaneously. The goal of this study is to develop high-throughput sequencing techniques to identify novel DNA markers allowing the two prognostic groups to be distinguished, informing subsequent treatment decisions.
Genome-scale Characterization of Chromosonal Aberrations Using Parallelizable Compression Algorithms.
Dan Simovici (UMass Boston), Nurit Haspel (UMass Boston), David Weisman (UMass Boston), and Jennifer Rosen (BU) are seeking to speed up detection of genetic mutations that cause cancer. One of the great difficulties in developing effective cancer treatments is the stochastic or statistically varying nature of these mutations – each patient and each tumor may present a unique genomic signature. As individual full-genome sequencing comes within reach, this research will develop new data compression techniques (essentially automatically locating and leveraging repetitive sequences) for large-scale data mining of next-generation DNA sequencing data to speed-up detection of large chromosomal aberrations.
Automated Segmentation of Vessel Network Structures in Large Image Stack Sets.
Deniz Erdogmus (Northeastern) Jeff Lichtman (Harvard) and David Boas (Harvard/Mass. General Hospital) are developing software that will improve our understanding of the complex behavior of neuronal circuits, and how they evolve during synaptic competition. Today neuroscientists must manually construct the tree-structured nerves from 3D microscopy imagery. Manual reconstruction of a single cell may take weeks to months depending on its type and complexity. In this project, the team seeks to develop an open high-performance automatic image analysis software package to massively accelerate this process, making it feasible to analyze automatically or interactively a large imagery database for quantitative modeling and statistical analysis.
Hossein Mosallaei and David Kaeli (Northeastern) have teamed up with Efthimios Kaxiras (Harvard) to create computer models that simulate the behavior of metals, dielectric and magnetic particles at extremely small scales, allowing insights into the behavior of important new materials.
Greening High Performance Computing
Ayse K. Coskun (BU) Martin C. Herbordt (BU), and Gunar Schirner (Northeastern) are exploring ways to measure and improve the operating cost and energy efficiency of large-scale high-performance computing systems by taking a holistic view of user demand, server power consumption, and facility infrastructure along with energy market features and constraints.
Massive Genomic Data Processing and Deep Analysis.
Yanlei Diao and Li-Jun Ma (UMass Amherst), together with Samuel Madden (MIT), and Yiping Shen and Bai-Lin Wu (Harvard Medical School & Children’s Hospital Boston) are developing a next-generation, on-demand service for managing and processing massive amounts of genome information and automating the analysis of association, causality, and outlier detection. MORE
Making matrix calculations 1000 times faster
Lorena Barba (Boston University), Cris Cecka (Harvard) and Hans Johnston (UMass Amherst) are developing computational techniques to accelerate matrix calculations (a fundamental building block for many High Performance Computing applications) in future generation “exascale” software platforms, which will be 1,000 times faster than current computing speeds.