Software Productivity and Sustainability for Science and Engineering, Anshu Dubey, Argonne National Laboratory


Computing has become a critical component of science and engineering. Software is the key crosscutting technology that enables advances in mathematics, computer science, and domain-specific science and engineering to achieve robust simulations and analysis for predictive science, engineering, and other research fields. While software is becoming more complex due to multiphysics and multiscale modeling, the coupling of data analytics, and disruptive changes in computer hardware (due to increases in typical system scale and heterogeneity, including GPUs and additional alternative architectures), software itself has not traditionally received focused attention in the CSE community or been rewarded by that community. This Guided Affinity Group will introduce topics in a minisymposterium and several minisymposia that address growing technical and social challenges in software productivity, quality, and sustainability, and thereby are helping software fulfill its critical role as a cornerstone of long-term CSE collaboration

What are the relevant conference themes?

  • Applications in science, engineering, and industry
  • Emerging software infrastructure for CSE, sustainability of numerical software
  • High-performance computing, emerging architectures and programming paradigms
  • Multiscale, multiphysics, and multilevel methods

Short Biography:

Anshu Dubey is a computer scientist in the Mathematics and Computer Science Division at Argonne National Laboratory and a Senior Scientist at the University of Chicago. She led the earth and space sciences sub-area of applications in the US-DOE Exascale Computing Project for two years. She is interested in all aspects of HPC scientific software including the numerics, design and especially productivity issues. She is the chief software architect for FLASH, a multiphysics HPC software that is used by multiple science and engineering domains.


I firmly believe that science and engineering done through computing is only as good as the software used in the process. Although there is growing awareness of software quality impact on CSE, investment in both process and people is lagging. I have long been an advocate for both, and also for more research in software engineering for exploratory software. I take up every opportunity for advancing the cause of software productivity and sustainability for CSE software in the hope that repeated exposure will cause the decision makers in the community to give this topic the attention it deserves. I believe that SIAM-CSE conferences have the audience that is critical for this purpose. Given my 15+ years of experience in leading the development of a multiphysics multi-domain community scientific software, I believe I am well positioned to lead this GAG.