Skip to main content

Software and Architecture for Reliable Quantum Computing

ECE General Seminars ECE Seminar

-
Location: EER 3.646
Speaker:
Poulami Das
Georgia Institute of Technology

Quantum computers can solve many important applications that are beyond the capabilities of conventional computers. However, noisy quantum hardware leads to errors during program execution which limit us from running applications with high fidelity. Quantum information can be protected by using quantum error correction, but it requires huge resource overheads that are impractical on near- term systems with a few hundreds of qubits. These near-term systems can speed up certain domain- specific applications even in the presence of errors. Simultaneously, they are used to understand error correction at a small-scale to pave the pathway toward fault-tolerant systems that can unlock the full potential of quantum computing.
 
In this talk, I will provide an overview of my contributions in the quantum software and architecture stack to improve the reliability of quantum computers. I will introduce ANGEL, a quantum compiler that leverages application and device error characteristics to reduce the impact of gate errors during program execution. ANGEL creates proxy circuits that imitate a given application, runs them on the device, and builds an error profile which is then used to learn the optimal sequence of low-level gates to
translate the program. Next, I will describe the challenges in enabling real-time quantum error correction within a budget of a few micro-seconds and discuss LILLIPUT, an efficient lookup table error decoder. LILLIPUT only stores entries corresponding to correctable errors and significantly reduces the memory overheads of the lookup tables. I will also briefly discuss the system-level organization required for quantum error correction at large scale. Finally, I will conclude with my future research vision towards building a unified software stack for quantum computing, designing systems for enabling scalable quantum computers, and architecting cryogenic systems.

Biography

Poulami Das is a PhD Candidate at Georgia Institute of Technology. Her research focuses on developing software and architecture for improving the reliability of quantum computers. She obtained her Masters degree from the University of Texas, Austin and Bachelors degree from the National Institute of Technology (NIT), Durgapur, India. She is the recipient of the Microsoft Research PhD Fellowship, Institute Gold Medals at NIT Durgapur, and has been selected as a Rising Star in EECS. She is passionate about continuing innovative research to pave the road towards practical quantum advantage and broadly impact the field of computing as well as the society at large. Her research has been recognized with the Best Research Award at the DAC PhD Forum, Cleaver Award for the most outstanding PhD dissertation proposal in ECE, Georgia Tech, a Best Paper Award at Computing Frontiers, and has appeared in top architecture and systems venues like MICRO, HPCA, and ASPLOS. She frequently collaborates with quantum and architecture research groups at IBM, Microsoft, Google, and Amazon. She is also passionate about teaching and most recently has served as a co-instructor for a graduate-level course on quantum computing at Georgia Tech.

Seminar Series