Singular is a computer algebra system for polynomial computations, with special emphasis on commutative and non-commutative algebra, algebraic geometry, and singularity theory. The goal of the project is to provide highly efficient algorithms to handle Gröbner resp. standard bases and free resolutions, polynomial factorization, resultants, characteristic sets, and numerical root finding. The project uses GPI-Space to enable massively parallel computations in computer algebra, allowing applications to scale from personal computers to thousands of cores on HPC clusters. So far, applications have appeared in the fields of algebraic geometry, tropical geometry, geometric invariant theory, high energy physics, and condensed matter physics. Singular relying on GPI-Space for state-of-the-art parallelization infrastructure also contributed new programming constructs back into GPI-Space.
Industrial Seismic Imaging
Seismic imaging is a demanding field. Data sizes in the order of dozens of Terabytes meet plenty of complex processing steps and optimization workflows. Strict quality measures meet short turnaround cycles and the constant request for real time analysis. Built on internal domain expertise and in close cooperation with industrial partners GPI-Space is used with great success as core technology in industrial Seismic Imaging providing scalable and robust infrastructure in highly dynamic environments and for very long running applications.
In cooperation with the Leibniz Computing Centre and the University observatory Munich GPI-Space is used as integration platform for the particle visualization tool SPLOTCH. Within 3 work weeks the end to end throughput of SPLOTCH was boosted by a factor of 10x. Crucial is the automatic advanced scheduling and communication hiding provided by GPI-Space.