/assets/IISc.jpeg

About

Main Building at IISc, Bangalore

Hi, my name’s Gourab and I’m a PhD Student working under supervision of Prof. Phani Motamarri at MATRIX Lab, Department of Computational and Data Sciences, Indian Institute of Science (IISc), Bangalore. My research focuses on developing hardware-aware computational algorithms and novel implementation techniques to accelerate large-scale iterative eigenvalue and linear solvers on heterogenous architectures. This includes developing matrix-free methods, mixed-precision strategies, compressed basis techniques and communication avoidance algorithms. These procedures can be applied to accelerate solvers employed in various fields, in particular, the quantum modeling of materials which are solved on extreme-scaling computing architectures using finite element discretization of Kohn-Sham density functional theory equations.

Education

M.S. in Physics (2018-2019)
Indian Institute of Science, Bangalore

Implementated a time-dependent density functional theory module in an in-house python port of quantum espresso and applying it to study the response or the induced polarization when a time-dependent electric field is applied to a system. Masters Thesis titled “Time-dependent Density Functional Theory” under Prof. Manish Jain.

B.Sc. (Research) in Physics (First Class) (2014-2018)
Indian Institute of Science, Bangalore

Studied Non-Hermitian PT-Symmetric Quantum Hamiltonian models used to describe nonequilibrium steady states of dissipative quantum systems. Bachelors Thesis titled “Orbital Effect of Magnetic Field on a Current Driven System” under Prof. Sumilan Banerjee.

Experience

Contributed to DFT-FE (Das et al., 2022)
(Winner of the 2023 ACM Gordon Bell Prize, the highest prize in High Performance Computing)

DFT-FE (C++/CUDA/MPI) is a massively parallel finite-element based open-source code for material modeling using density functional theory (DFT). DFT-FE can handle periodic, non-periodic and semi-periodic calculations. It can further handle both pseudopotential (ONCV, Troullier-Martins) and all-electron (non-relativistic) calculations; XC functionals supported are LDA, GGA (including spin-polarized); Capabilities include ground-state calculations, structural relaxations (ionic, cell), ab-initio MD (NVE). Scalability tested on ~200,000 MPI tasks on many-core architectures and ~64,000 GPUs on hybrid CPU-GPU architectures.

Project Associate (2020-2021)
MATRIX Lab, Department of Computational and Data Sciences, IISc

Developed hardware-aware computational algorithms based on the matrix-free paradigm to evaluate on-the-fly finite-element discretized matrix-multivector products arising in solving the DFT problem without storing the cell-level matrices or the global sparse matrix under supervision of Prof. Phani Motamarri.

KVPY Summer Project (May 2016 - July 2016)
Theoretical Condensed Matter Group, Department of Physics, IISc

Studied Ising model and Spin systems using Neural Networks under Prof. Tanmoy Das

KVPY Summer Project (May 2015 - July 2015)
Tomography Lab, Department of Physics, IISc

Studied the variation in DOLP of light propagating through turbid media under Prof. Rajan Kanhirodan

Treasurer (2024 - 2025)
IISc ACM Student Chapter

Handled the finances of and organized various events by IISc ACM Student Chapter

Honors and Awards

ACM Gordon Bell Prize
Issued by ACM (The highest prize in High Performance Computing)

Prime Minister’s Research Fellowship (PMRF)
Issued by Ministry of Education (MoE), Government of India

IEEE TCHPC Travel Award
Issued by IEEE

Kishore Vaigyanik Protsahan Yojana Fellowship (KVPY)
Issued by Department of Science and Technology (DST), Government of India

Skills

  • Expertise in low-level kernel programming using C/C++, CUDA, HIP and SYCL, specializing in developing high-performance, scalable codes leveraging OpenMP and MPI for parallel computing.

  • Proficient in Python for data analysis and algorithm development and developing dynamic web applications using Svelte.

Teaching Assistant


Copyright © 2025 Gourab Panigrahi
Powered by Urara