Sparse Solvers for Exascale: From Building Blocks to Applications



The workshop focuses on the issues of high-performance and exascale computing from the perspective of building blocks, algorithms, and applications. Specific topics addressed in the contributions include
  • Fault tolerance
  • Energy awareness
  • Sparse Solvers
  • Krylov, Jacobi-Davidson, Chebyshev techniques
  • Sparse matrix kernels
  • Hybrid programming
  • Computation of eigenstates and spectral information
  • Time evolution
  • Quantum physics at the exascale level

The principal intentions of the workshop are similar to those of the DFG priority programme "Software for Exascale Computing (SPPEXA)".


Invited Talks

Edmond Chow
“Quantum Chemistry on Large-scale Heterogeneous Clusters”
Marlis Hochbruck
“Time integration of evolution equations”
Satoshi Matsuoka
“Harnessing the Deep Memory Hierarchy with Communication Reducing Algorithms for Exascale”
Yousef Saad
“Divide and conquer algorithms for large Hermitian eigenvalue problems”
Horst Simon
“Usable Exascale and beyond Moore’s Law”

Public Evening Lecture (Tuesday)

Horst Simon
“What Supercomputers Do, and what Supercomputers Still Can't Do”


The workshop begins on Monday morning, 9.00am, and ends on Wednesday afternoon, about 17.00pm.

A welcome reception (food+drinks) is scheduled for Sunday evening, 18.00pm, and refreshments will be offered after the last session on Wednesday.

Final details and the precise schedule with all talks will follow shortly.

Local registration
Sunday, 17.00-19.00 in the Alfried-Krupp-Wissenschaftskolleg, or during the workshop
Welcome reception
Sunday, from 18.00 in the “Brasserie Hermann”
Workshop opening & first session
Monday, 9.00
Workshop summary & farewell
Wednesday, 17.00