Scientific Computing

scientific computingSimulation plays a major role in nearly every area of science and engineering—from data analysis to physical models. Our faculty design, build, and analyze the behavior of numerical algorithms to ensure that numerical methods are accurate and that implementations are efficient.

We design and analyze the accuracy of methods, developing numerical approximations to partial differential equations with advanced finite element methods and integral equations. We also develop solvers for these problems, instrumenting techniques based on numerical linear algebra, iterative subspace methods, and multigrid methods. Our research explores the efficiency of these methods on a range of architectures and environments, from high-concurrency nodes, such as GPUs, to large-scale supercomputing systems. We explore parallel scalability and analyze performance in computing kernels from graph algorithms to sparse linear algebra.

CS Faculty and Their Research Interests

Paul Fischer numerical PDEs, spectral element methods, computational fluid dynamics, parallel and high-performance algorithms, iterative methods 
William Gropp high performance scientific computing, scalable numerical algorithms for PDEs, large-scale parallel software 
Michael T. Heath numerical analysis and scientific computing, numerical linear algebra and optimization 
Laxmikant Kale simulation software, numerical libraries and algorithms 
Andreas Kloeckner integral equation methods for PDEs, high-order finite element methods for hyperbolic PDEs, tools and languages for high-performance computing, time integration 
William Kramer extreme-scale computing and analytics, performance evaluation, data and storage techniques 
Luke Olson numerical analysis, scientific computing, large-scale simulation 
Marc Snir large-scale parallel systems, algorithms, and libraries 
Edgar Solomonik communication complexity

Affiliate Faculty

Robert Brunner, Astronomy computational astrophysics
Daniel S. Katz, NCSA resilience and fault-tolerance, many-task computing, parallel and distributed computing, sustainable and open science software

Adjunct Faculty

Frank Cappello,
Argonne National Lab
determinism in high-performance and distributed computing, check-pointing, fault prediction 

Related Scientific Computing Research Efforts and Groups

Scientific Computing Research News

Thomas M. Siebel Chair in Computer Science and National Center for Supercomputing Applications Professor Bill Gropp

University Supercomputers Are Science's Unsung Heroes, And The Fastest Yet Is On The Way

August 30, 2018  

Popular Science -- “The machines are all over subscribed,” Bill Gropp, Illinois CS Professor and director of the National Center for Supercomputing Applications, says as work starts for what will be the next, fastest supercomputer, Frontera. Illinois' Blue Waters has been used to do things like model an enormous EF-5 tornado and to produce maps of Alaska.


Professor Paul Fischer has won an $800,000 grant from the Department of Energy.

Fischer Plans to Use Grant to Improve Accuracy, Reduce Cost of Fluid-Flow Models for Nuclear Reactors

July 17, 2018   An $800,000 Department of Energy grant will support work to create fluid-flow models for nuclear reactors that can run on desktop computers.
CS Assistant Professor Jian Peng

Satellites, Supercomputers, And Machine Learning Provide Real-time Crop-Type Data

April 4, 2018 -- Research from Illinois that includes work by CS Assistant Professor Jian Peng has provided a new technique for distinguishing between corn and soybeans in satellite data, crops that previously have been impossible to distinguish from space. Also covered by Farmers Advance.

Kale, Zhai Named ACM Fellows

Kale, Zhai Named ACM Fellows

January 18, 2018   Professors Lamikant "Sanjay" Kale and ChengXiang Zhai were among 54 new ACM Fellows named for 2017.
CS Assistant Professor Andreas Kloeckner

Kloeckner wins NSF CAREER Award to make computer simulation faster, cheaper

September 20, 2017   Kloeckner looks to create better, more versatile solvers for elliptic PDEs in order to improve large-scale simulations.