Abstract
Quantum computers promise transformative impact by enabling us to run programs that are beyond the scope of even the most powerful supercomputers. However, noisy quantum hardware limits us from running most practical quantum algorithms because program execution encounters frequent errors. Professor Das’ research focuses on developing software and architectural mechanisms to improve the fidelity of quantum computers.
In this talk, Professor Das will discuss compilation for newer generations of quantum hardware employing distributed architectures. These machines connect high-fidelity small-sized chips via interconnects to scale system sizes but rely on error-prone and slow teleportations to execute operations between qubits on different chips. She will discuss compilation methods to generate optimized program schedules with reduced vulnerability to teleportation errors. Next, she will describe fine-grained instruction scheduling to improve the performance of fault-tolerant quantum computers. These systems spend most of their time running quantum error correction (QEC) cycles to detect and correct hardware errors. The length of QEC cycles dictates error resilience and program runtime. She will discuss hardware-software mechanisms that shorten these cycles by dynamically adapting QEC instruction latencies based on error events observed in real-time to optimize system performance.
Biography
Poulami Das is an assistant professor in the department of electrical and computer engineering at the University of Texas at Austin (UT Austin). Her research interests include quantum computing, computer architecture, and systems for efficient large language model (LLM) inference. Her research has been recognized with several awards, including the NSF CAREER Award, UT ECE Junior Faculty Spirit Award, and has appeared in top architecture, machine learning, and systems conferences, such as MICRO, HPCA, ISCA, ASPLOS, EuroSys, MLSys, ICML, and NeurIPS. She received her doctoral degree from Georgia Tech and master’s degree from UT Austin.
