Resume Michal Merta

Michal Merta
computational scientist in HPC / applied mathematician

Researcher at IT4Innovations National Supercomputing Center. He is mainly interested in high performance computing and the parallel boundary element method (BEM). He is a co-founder of the library BEM4I that aims at parallel solution of problems from linear elasticity to sound scattering using BEM. Currently he focuses on an acceleration using the Intel® Xeon Phi™ coprocessors. He was awarded Joseph Fourier Prize 2015 by Bull s.r.o. and Embassy of the French Republic, and Babuska Prize 2011 by Czech Society for Mechanics for his research.

Work experience

2015 - present

Researcher
IT4Innovations National Supercomputing Center, Ostrava, Czech Republic

Acceleration of numerical libraries using Intel Xeon Phi coprocessors. Development of the boundary element method based solvers.

2015

Junior Researcher Assistant
IT4Innovations National Supercomputing Center, Ostrava, Czech Republic

Parallelization and acceleration of the boundary element method based solvers.

2011 - 2014

Research Assistant
IT4Innovations National Supercomputing Center, Ostrava, Czech Republic

Development of the boundary element method based solvers.

2011 - 2014

Teaching
VSB - Technical University of Ostrava, Ostrava, Czech Republic

Mathematical analysis I., II.

Education

2009 - 2011

MSc. in Computational Mathematics
VSB - Technical University of Ostrava, Ostrava, Czech Republic

Diploma thesis: Discretization and numerical realization of the dynamical contact problems with friction.

2006 - 2009

BSc. in Computational Mathematics
VSB - Technical University of Ostrava, Ostrava, Czech Republic

Diplomat thesis: Newmark method for the solution of initial problems of the second order.

Research interests

High performance computing, Intel® Xeon Phi™accelerators, boundary element method.

Projects

ipcc.it4i.cz

Intel Parallel Computing Center at IT4Innovations
Acceleration of the ESPRESO library using Intel® Xeon Phi™ coprocessors

Main activities of the Intel® PCC at IT4I are divided into two pillars: The Development pillar of highly parallel algorithms and libraries focuses on the development of the state-of-the-art sparse linear iterative solvers combined with appropriate preconditioners and domain decomposition methods, suitable for solutions of very large problems distributed over tens of thousands of Intel® Xeon Phi™ coprocessors accelerated nodes. Developing solvers will become part of the IT4I in-house ESPRESO (ExaScale PaRallel FETI SOlver) library. The support of HPC community codes includes creating interface between ESPRESO and existing community codes Elmer and OpenFOAM Extend Project.

Software Development

2012 - present

BEM4I
Parallel Boundary Element Solver

BEM4I aims at parallel solution of problems from linear elasticity to sound scattering using the boundary element method. The library leverages the techniques of matrix sparsification (ACA, FMM), is parallelized in shared and distributed memory, and features explicit vectorization. Its acceleration using Intel® Xeon Phi™ is currently being implemented.

2015 - present

ESPRESO
ExaScale PaRallel FETI SOlver

ESPRESO (ExaScale PaRallel FETI SOlver) is a sparse iterative solver based on the Finite Element Tearing and Interconnect (FETI) methods. Solver uses the Hybrid FETI method based on a multi-level decomposition which significantly improves the scalability to the tens of thousands of compute nodes solving tens of billions of unknowns. ESPRESO also supports both Nvidia GPU and Intel Xeon Phi accelerators which bring significant speed up for problems requiring high number of iterations.

Teaching and Lecturing Experience

2011 - 2014

Mathematical Analysis I., II.
Undergraduate level course, VSB - Technical University of Ostrava, Ostrava, Czech Republic

2012

Introduction to Scientific Computing Using PETSc and Trilinos.
PRACE Spring School, Krakow, Poland