Midwest Research Computing and Data Consortium
In this series, we dive into the world of the Midwest Research Computing and Data Consortium, exploring its members, their challenges, and future prospects. Recently, we had the opportunity to speak with Doosoo Yoon, a Research Computational Facilitator at the University of Iowa, who shared his insights on the field. Edited excerpts of our conversation below:
As a Research Computational Facilitator at the University of Iowa, my role has two primary components. First, I introduce new users to our advanced computing facility through workshops and hands-on activities. Second, I assist users in resolving any issues or concerns they encounter with our cluster. We have a team of experts in various fields such as programming, cyberinfrastructure, storage, and interactive platforms, and we work together to provide reliable solutions to our users. My role also includes consultations with users about applying for large-scale projects to nationwide supercomputing facilities like the ACCESS program.
Having a background in computational astrophysics and data science helps in that I have had the opportunity to meaningfully engage in international collaborations, which has in turn significantly influenced my current role. Working as a computational astrophysicist taught me the value of using large-scale simulations and collaborating with experts from various fields. My involvement in the Event Horizon Telescope collaboration, which included over 250 international researchers, highlighted the importance of building a supportive community and working with diverse teams.
Machine learning is becoming increasingly integral to high-performance computing. My experience with machine learning includes integrating it into hydrodynamic simulations to make computations more efficient. At the University of Iowa, we provide resources such as GPUs and programming libraries like PyTorch and TensorFlow to support machine learning projects. Our team also works closely with users to help them utilize these resources effectively.
My approach to facilitating the needs of researchers from various disciplines involves active listening and understanding their requirements. In this role, you need to build connections, and having genuine interest across different research areas that require specific input(s) helps a lot. I open projects with them, get as involved as possible, and keep communicating through the course of the project. This makes it easier to provide technical support and resolve any issues they encounter.
One of the biggest challenges in research computing is the rapid growth in demand for computing services. Many people are still unaware of the available facilities. Advertising these resources and encouraging community use is crucial. Networks like the Midwest RCD can help by fostering a collaborative environment and providing necessary support.
Looking towards the future, I see Artificial Intelligence (AI) having a significant impact on high-performance computing. There’s a lot of discussion about how to use AI effectively and ethically. The community is still evaluating the best ways to integrate AI into high-performance computing and we are likely to see substantial advancements in this area in the coming days.
For students or professionals from non-traditional backgrounds looking to enter the field of research computing, my advice is not to be afraid to try new things. Consult with research computing facilitators if they are available in your community. Learning everything on your own can be overwhelming so reach out for help and guidance.
Similar to work-life balance, it is a common difficulty for researchers to balance the timing required between conducting research and obtaining new technology. Therefore, a strategic approach for this balance is critical for research productivity, sustainability, and innovations. Understanding your goals before diving into high-performance computing can save a lot of time and effort.