ATS Seminar Series: Wesley Brewer

The Advanced Technologies Section (ATS) of the National Center for Computational Sciences at ORNL is a world leader in developing and deploying scientific and technical solutions for leadership-class computing environments. The R&D activities of ATS are organized around designing and deploying leadership class systems, developing artificial intelligence solutions for science and smart facilities of the future, and stewardship of data and workflows at scale to enable science. The ATS Seminar Series is a forum for learning from experts and engaging with collaborators to advance their scientific mission.


Seminar Title:

Production Deployment of Machine-Learned Surrogate Models on HPC

We explore how to optimally deploy several different types of machine-learned surrogate models used in rotorcraft aerodynamics on HPC. We first developed three different rotorcraft models at three different orders of magnitude (2M, 44M, and 212M trainable parameters) to use as test models. Then we developed a benchmark, which we call “smiBench”, that uses synthetic data to test a wide range of alternative configurations to study optimal deployment scenarios. We discovered several different types of optimal deployment scenarios depending on the model size and inference frequency. For most cases, it makes sense to use multiple inference servers, each bound to a GPU with a load balancer distributing the requests across multiple GPUs. We tested three different types of inference server deployments: (1) a custom Flask-based HTTP inference server, (2) TensorFlow Serving with gRPC protocol, and (3) RedisAI server with SmartRedis clients using the RESP protocol. We also tested three different types of load balancing techniques for multi-GPU inferencing: (1) Python concurrent.futures thread pool, (2) HAProxy, and (3) mpi4py. We investigated deployments on both DoD HPCMP’s SCOUT and DoE OLCF’s Summit POWER9 supercomputers, demonstrated the ability to inference a million samples per second using 192 GPUs, and studied multiple scenarios on both Nvidia T4 and V100 GPUs. Moreover, we studied a range of concurrency levels, both on the client-side and the server-side, and provide optimal configuration advice based on the type of deployment.



Wesley Brewer

Computational Scientist / Engineer

Wesley Brewer is a contractor with the Department of Defense’s High Performance Computing Modernization Program (HPCMP), working as a computational scientist/engineer on the User Productivity and Enhanced Training (PET) team. His current research is in the area that intersects computational fluid dynamics (CFD), artificial intelligence, and supercomputing, in which he has published numerous scientific papers. In addition to these areas, he has also published in the areas of numerical weather prediction, computational genetics, and cloud computing. He has won a number of awards for his research, including the 2019 DoD HPCMP hero award for technical excellence. He has a B.S. in Engineering Science & Mechanics from the University of Tennessee, an M.S. in Ocean Engineering from the Massachusetts Institute of Technology, and a Ph.D. in Computational Engineering from Mississippi State University.

Dial-in Information


Microsoft Teams meeting

Join on your computer or mobile app

Click here to join the meeting

Or call in (audio only)

+1 865-276-6990,,324719724#   United States, Knoxville

Phone Conference ID: 324 719 724#

Find a local number | Reset PIN

Learn More | Meeting options



Dec 13 2021


Eastern Standard Time, UTC-5
12:00 pm - 1:00 pm




Arjun Shankar
(865) 574-2704


QR Code

Comments are closed.