NVIDIA GPU Award for Best GPU Poster

Posted on

Tim Duong

Tim Duong, an undergraduate researcher in the Neuhauser group, received the Best GPU Poster Award during the COMP Poster Session at the NVIDIA GPU Award Symposium at the recent American Chemical Society (ACS) Fall 2023 meeting.

The NVIDIA GPU Award for Best GPU Poster recognizes exceptional computational chemistry research using graphics processing units (GPUs), including developing accelerated algorithms, conducting simulations with GPU-accelerated software (e.g., AMBER, NAMD, GROMACS, etc.), or a combination of both. Duong was awarded a high-performance workstation GPU. Click here to view a PDF of Duong’s award-winning poster.

Undergrad researcher Tim Duong (Neuhauser group) proudly presents his award-winning poster with the accompanying high-performance workstation GPU prize.

“Our research project focuses on the ability of GPUs and its superior parallelization compared to central processing units (CPUs) to accelerate the recently-developed stochastic GW (sGW) method,” Duong explained. “Many quantities in the sGW method such as our stochastic orbitals are independent of one other. As a result, calculations such as the time propagation of these orbitals can be run separately and in parallel. This makes sGW much more computationally efficient and advantageous when run on GPUs rather than CPUs. Our GPU-implemented sGW specifically uses NVIDIA CUDA technology and NVIDIA A100 Tensor Core GPUs for the time-propagation of our stochastic orbitals. We plan to demonstrate this method by calculating the quasiparticle energies of systems such as chromophores and silicon clusters of varying sizes that contain thousands of atoms. These systems are typically computationally inaccessible with conventional quantum mechanical (QM) methods due to the system sizes, but they are within reach and computationally affordable when using GPU-implemented sGW. This development will provide a new method of computation in computational chemistry that takes advantage of the strengths of GPUs and make QM-level calculations of very large systems not only possible but also highly efficient.”

In his 2nd year as a chemistry major, Duong specializes in theoretical chemistry. He conducts research in Professor Daniel Neuhauser’s group, exploring stochastic electronic structure theory and quantum dynamics. His focus is on calculating quasiparticle energies for extensive systems. The UC LEADS scholar intends to pursue a theoretical chemistry Ph.D. after graduation.

The award is sponsored by the NVIDIA Corporation, an American technology company that specializes in designing and manufacturing GPUs, as well as other related hardware and software products. NVIDIA’s GPUs are widely used in various fields including gaming, artificial intelligence, data science, and professional graphics applications.

Penny Jennings, UCLA Department of Chemistry & Biochemistry, penny@chem.ucla.edu.