Google Summer of Code
Summer, 2023
Tuitus Foundation participated in Google Summer of Code through NumFOCUS in 2023. We supported 1 undergraduate student working on developing a Differentiable Material Point Method. For information on the project refere to the blog post.
NHERI Computational Academy
July 17 - 20, 2023, TACC, UT Austin
In collaboration with DesignSafe-CI and SimCenter, Tuitus Foundation organized a hackathon for 40 participants working on natural hazards. Two out of the eight teams were developing hacks on the open source software tools supported by the Tuitus Foundation. The winning team using Adcirc to model flooding damage and its impact on the community.
For more information, please see NHERI Academy website
PEARC Workshop Tutorials - Best Practices of CI/CD for High Performance Computation with Tapis Workflows API
July 23 - 27, 2023, Portland, Oregon
We organized a tutorial at the PEARC 2023 conference, the tutorial session was attended by 28 participants. The tutorial session covered a range of topics related to developing research computing workflows. We delved into the challenges researchers face, particularly the need to understand DevOps tooling and software design patterns that often fall outside their domain expertise. The session introduced the Tapis Workflows API, a production-grade service that aids researchers in simplifying their workflow creation process.
The API’s strength lies in its ability to abstract the complexities of underlying technologies, presenting them through a user-friendly interface. This interface seamlessly integrates with HPC resources at any institution equipped with a Tapis deployment. We discussed how the Tapis Workflows API empowers users to design and automate workflows. These workflows can build container images, run scientific computing jobs, and even trigger subsequent tasks on independent systems using webhook notifications. A figure was presented to visually demonstrate this process.
A significant portion of the tutorial was dedicated to the NSF-funded Tapis v3 platform. This Application Program Interface (API) is designed for distributed computation, and we explored how it can be used to create scientific workflows effectively. Techniques were shared that would enable attendees to automate and validate their CI/CD pipelines for scientific computing applications.
One of the highlights of the session was the hands-on exercises. These exercises were designed to provide attendees with practical experience in building an automated CI/CD pipeline for a large scientific application. We emphasized the flexibility of these workflows, highlighting how they can be transitioned between different execution environments, from small virtual machines to national-scale supercomputers. Overall, the attendees left with a comprehensive understanding of the Tapis Workflows API and its potential in revolutionizing research computing workflows.
