Thread: 2022 SSA Session: From Desktops to HPC & Cloud: Emerging Strategies in Large-scale Geophysical Data Analysis

Started: 2022-01-05 15:21:07
Last activity: 2022-01-05 15:21:07
Topics: SSA Meetings
Dear Colleagues,

Are you working with large-scale geophysical data sets? Please consider sharing your approaches and results, whether they were completed on a Raspberry Pi farm, GPU cluster, HPC resource, or a cloud environment in the session below. We are looking forward to discussing your work, hopefully in person, this April!

Abstracts are due by January 12th (next week):

Session: From Desktops to HPC & Cloud: Emerging Strategies in Large-scale Geophysical Data Analysis

As the availability of geophysical data continues to grow in volume and variety, many aspects of research data collection, access and processing are evolving to allow full use of large data sets. Processing large data volumes is not unique to geophysics and there exist many modern, open source languages (e.g. Python, Julia), data containers (e.g. HDF5, Zarr) and computational frameworks (e.g. Apache Spark, xarray and Dask, Ray), that can be leveraged and allow researchers to focus more on the domain-specific issues. Access to computational resources, such as HPC and cloud computing, continue to become more accessible and affordable. Specialized hardware, such as GPUs, are increasingly available in both academic and commercial computing environments and make efforts such as large scale waveform template matching possible. New computing models, like serverless architectures and Kubernetes container orchestration, expand the ways in which research can be performed. The combination of available software and computational resources increase accessibility to a new scale of inquiry, making large-scale research in seismology, infrasound, geodesy and geophysics in general more tractable than ever before. In this session, we invite researchers, data producers and data providers to share work in data-hungry applications, approaches to large data collection, storage and access and experiences with processing platforms and architectures.

Chad Trabant, Incorporated Research Institutions for Seismology (chad.trabant<at>
Jonathan K. MacCarthy, Los Alamos National Laboratory (jkmacc<at>
09:05:20 v.01697673