Vanguards of HPC-AI: Oak Ridge Lab’s Verónica G. Melesse Vergara — Unleashing ‘Breakthrough Science’

Verónica G. Melesse Vergara

In our Vanguards of HPC-AI series, we now feature Verónica G. Melesse Vergara, Section Head, HPC Operations in the National Center for Computational Sciences Division at Oak Ridge National Laboratory.

She became involved in scientific computing in 2008 as an undergraduate at Reed College in Portland, and her involvement in HPC began in 2009 at Florida State University.

Melesse Vergara is from Quito, Ecuador. She earned a B.A. in Mathematics/Physics at Reed College and an M.S. in Computational Science at Florida State University. She has more than a decade of experience in HPC, having previously been Group Leader of the System Acceptance & User Environment Group in the National Center for Computational Sciences (NCCS), which is home to the Oak Ridge Leadership Computing Facility (OLCF).

She also is part of OLCF’s systems testing team and led acceptance for the Summit supercomputer (formerly the world’s no. 1 system) and for Frontier, housed at the OLCF and the world’s first certified exascale-class supercomputer (currently the world’s no. 2 system).

Her research interests include HPC, large-scale system testing, and performance evaluation and optimization of scientific applications. She is a member of IEEE and ACM and serves in the ACM SIGHPC Executive Committee and the SC Conference Steering Committee.

We congratulate ORNL’s Verónica G. Melesse Vergara on being recognized as an HPC-AI Vanguard.

HPC-AI Vanguards: Current and Future Leaders
Verónica G. Melesse Vergara
Section Head
HPC Operations in the National Center for Computational Sciences Division
Oak Ridge National Laboratory

What was your first involvement in HPC or AI?

I became involved in scientific computing in 2008 writing my senior thesis at Reed College in Portland. My professor and thesis advisor, Dr. Joel Franklin in the Physics Department, had a small Apple cluster that I used to run simulations of snowflake formation.

I got into HPC in 2009 as a graduate student at Florida State University working in the HPC Center at FSU as a Technical Support Assistant in lieu of a TA/RA-ship.

What is your passion related to your career path?

Solving problems! One of the most fun aspects of HPC engineering is that you get to solve a broad range of problems getting an application to run well and run at scale – sometimes to even run at all. There is always something new to learn and the feeling of finding a solution to a problem that unblocks and/or unleashes breakthrough science is extremely rewarding. Enabling and supporting a wide range of science domains is something very unique about our field.

Do you prefer working as an individual contributor or a team leader?

It depends on the specific project. I enjoy making individual technical contributions to larger projects to meet ambitious goals. It is how I have learned about a wide range of topics that I may not have been able to learn without hands-on practical experience. As I’ve advanced in my profession and gained more experience, most of my work revolves around leading small and large teams to accomplish ambitious goals. It is a very rewarding role and allows me to work with individual team members to identify opportunities for their own professional growth. Seeing how much more impactful a project is when it is the result of a collaborative effort that includes broad perspectives has shaped my role and attitude as a team leader.

Who or what has influenced you the most to help you advance your career path in this community?

The “who” is difficult to answer succinctly. Many folks have helped me along the way – from my supervisor at FSU for suggesting I apply for the Student Volunteer program at SC10, to my group leader at Purdue for supporting my involvement in the community, and many folks at ORNL and in the HPC community that I met through SC for encouraging me to pursue bigger and larger challenges and believing in me.

The “what” is easy: the SC conference and its community.

Your thoughts on how we, the nation, build a stronger and deeper pipeline of talented and passionate HPC and AI professionals?

Investing in programs that can bring a diverse set of folks to the field at multiple stages of the pipeline, with a special focus on the beginning of the pipeline. There are several efforts happening around the world, and finding ways to collaborate and capture those efforts would be useful. In the US, opening more avenues for folks that are interested but may not have experience yet – that would be great. The US has several programs but is often limited by citizenship, GPA, and focus areas, which can leave many potential candidates out.

What does it take to be an effective leader in HPC and AI?

Being willing to learn both from the continuously evolving landscape, but also engaging with the community. Finding ways to collaborate across organizations and fields is key for any effective leader in my opinion.

What is the biggest challenge you face in your current role? 

Now that we’ve entered the exascale era which has also intersected with the generative AI explosion, my biggest challenge, and in some sense opportunity, is to understand how the facility services and support ecosystem needs to evolve to adapt to the new requirements of our users.

What changes do you see for the HPC / AI community in the next five-10 years?  Technology?  Skill levels?  New application areas?  And how do you see your own skills evolving during this time frame?

A few things, including: a heavier focus on energy efficiency, emphasis on secure and trustworthy AI, growth in digital twins across domains, including climate, fusion, biology, and integration of HPC and quantum. As for myself, I’d like to dive deeper into AI-assisted modeling and simulation, and quantum computing and its applications in science.

Do you believe science drives technology or technology drives science?

Yes, absolutely. Scientific breakthroughs have made possible the devices we use on a regular basis today. And conversely, technology can drive science as we have seen in the last couple of years with the explosion of large language models and its applications. That boom has driven the market to develop chips optimized for that unique workload, in some cases at the expense of features used by today’s scientific applications.

Please share with us your personal interests and hobbies.

With husband Michael and friends

My husband Michael and I enjoy local breweries with our two “bears”, Ziggy and Zoe, who we adopted from nearby rescues during the pandemic. I also enjoy traveling, baking, and binge watching all kinds of series from sitcoms and dramas to telenovelas, docuseries, and reality TV.