Join the ChEESE Advanced Training on HPC for Computational Seismology
24 August 2020

Are you fascinated by high-performance computing and seismology? Are you curious to know how to numerically solve complex wave propagation problems?

Spend three full days interacting and learning from the developers of the ChEESE flagship codes for computational seismology completely free and online!

Save the dates from 21- 23 October 2020 for the ChEESE Advanced Training on HPC for Computational Seismology. You will have the opportunity to learn about the four HPC-powered software packages ExaHyPE, Salvus, SeisSol and SPECFEM3D.

ExaHyPE is a hyperbolic PDE engine used for solving systems of conservation laws to model seismic wave propagation problems.

Salvus uses a spectral-element approach to solve wave propagation problems and perform full-waveform inversion on unstructured meshes in 2 and 3 dimensions.

SeisSol solves seismic wave propagation and dynamic rupture problems using higher-order Discontinuous Galerkin (DG) methods on unstructured tetrahedral meshes.

SPECFEM3D is based on the higher-order spectral-element (CG) method for unstructured meshes. It can solve the linear seismic wave propagation in heterogeneous 3D models.

All attendees will have the opportunity to present their work in a lightning-talk session.

Click here for more information.

ChEESE infrastructures and datasets set to be integrated into EUDAT CDI
03 August 2020
ChEESE and EUDAT banner

The objectives and challenges of ChEESE are not only related to the exploitation of the enormous potential of the next generation of pre-exascale and exascale computing machines. There is an additional challenge, hidden somehow, but very much present to those who perform the day-to-day research activity, that is the effective and efficient management of the data to achieve the ambitious goals of the project.

ChEESE tackles the problem of having to work with a large amount of data in one of its work packages, together with the workflow management activity, with the concrete goal of extending and consolidating the use of EUDAT CDI data services among the partners. The choice to use EUDAT CDI data services for Solid Earth was pioneered by EPOS with a pilot project, but with the intention of extending the effort to other EPOS domains and services.

EUDAT CDI offers a wide range of generic data services, helping the research data management under different aspects: improvement on data preservation, data discoverability, reuse and accessibility, guarantees for failsafe and transparent data access, full identification, citation, the traceability of data and the reproducibility of (scientific) products; tools for facilitating data movement and analysis of large volumes of data, in line with the commitment of the ChEESE Consortium towards the completion of Open Science FAIR principles.

The working plan of the ChEESE data management group includes the identification of infrastructures and large datasets used and shared across the project, to collect their needs of data services through the whole project life and beyond. The actions aim at providing EUDAT data services to partners, as solutions for ChEESE Pilots Demonstrators data access, storage, and preservation, for the shared management of very large sets of data, processing, and analysis. The services chosen for storage and preservation of large datasets are B2SHARE/B2SAFE, while B2SHARE/B2NOTE are under consideration for metadata management and file annotation.

BSC and CINECA are CHEESE partners that will host two of the three European pre-exascale machines, and at the same time, they are EUDAT CDI members and EUDAT service providers.