Jason Chadwick
Jason D. Chadwick
Quantum computing Ph.D. student
University of Chicago

Google Scholar


I am a second-year computer science Ph.D. student at the University of Chicago studying quantum computer systems with Fred Chong. Previously, I graduated from Carnegie Mellon University with a B.S. in physics and a minor in computer science.

Hardware-software co-design is essential in quantum computing, both to get the best performance out of a given quantum device and to effectively guide the next generation of hardware. My research currently focuses on methods to improve the performance of quantum error correction on real quantum devices. I have most recently been working on software mitigation of time-varying noise (such as cosmic ray impacts and shifting two-level system defects) in the surface code. Previously, I have also published work in quantum control pulse optimization and a pair of papers on compiling with ququarts. My work so far has mostly had superconducting hardware in mind, but I have also worked with trapped ion and neutral atom architectures. I am currently working on moving more into the quantum error correction space, focusing on topics such as biased-noise QEC and decoding.


Averting multi-qubit burst errors in surface code magic state factories
under review
Jason D. Chadwick, Christopher Kang, Joshua Viszlai, Sophia Fuhui Lin, and Frederic T. Chong
[project]   [.pdf]   [arXiv]   [code]  
We design an efficient method to avoid cosmic ray errors in magic state factories, significantly reducing the qubitcycle cost of mitigating these errors.
Verity: a resilient kernel for magic state distillation
under review
Jason D. Chadwick*, Christopher Kang*, Sophia Fuhui Lin, and Frederic T. Chong
[project]   [slides]  
We show that surface code magic state factories are vulnerable to varying hardware performance and can magnify hardware degradation. To alleviate this problem, we introduce Verity, a resilient kernel for magic state distillation.
* indicates equal contribution
QTEM Best Paper - 3rd place
Efficient control pulses for continuous quantum gate families through coordinated re-optimization
QCE 2023
Jason D. Chadwick and Frederic T. Chong
[project]   [.pdf]   [publication]   [arXiv]   [poster]   [code]  
We present a method that allows quantum hardware to execute arbitrary operations at the pulse level by interpolating between a small number of known reference pulses. We demonstrate the procedure on the continuous space of all two-qubit operations.
Dancing the Quantum Waltz: Compiling Three-Qubit Gates on Four Level Architectures
ISCA 2023
Andrew Litteken, Lennart Maximilian Seifert, Jason D. Chadwick, Natalia Nottingham, Tanay Roy, Ziqian Li, David Schuster, Jonathan M. Baker, and Frederic T. Chong
[project]   [.pdf]   [publication]   [arXiv]  
We extend our previous work on qubit-to-ququart compression to specifically optimize three-qubit gates such as the Toffoli gate. We also find significant advantages in using Z-type multi-bit operations instead of X-type operations.
Qompress: Efficient Compilation for Ququarts Exploiting Partial and Mixed Radix Operations for Communication Reduction
Andrew Litteken, Lennart Maximilian Seifert, Jason D. Chadwick, Natalia Nottingham, Jonathan M. Baker, and Frederic T. Chong
[project]   [.pdf]   [publication]   [arXiv]  
We consider selectively compressing pairs of qubits into single four-state ququarts. We generate efficient "partial" operations between ququarts and qubits, which motivates a compiler that can transform any quantum circuit into this framework.
Time-Efficient Qudit Gates through Incremental Pulse Re-seeding
QCE 2022
Lennart Maximilian Seifert*, Jason D. Chadwick*, Andrew Litteken, Frederic T. Chong, and Jonathan M. Baker
[project]   [.pdf]   [publication]   [arXiv]   [poster]  
We present a method to iteratively obtain short-duration quantum control pulses when it is not possible to directly modify the objective function. We use this to find gate durations for high-radix logic gates that scale better than expected.
* indicates equal contribution
Prediction of electron density and pressure profile shapes on NSTX-U using neural networks
Nuclear Fusion 61 046024
Mark D. Boyer and Jason D. Chadwick
[project]   [.pdf]   [publication]   [poster]   [slides]  
We develop a neural network to accurately predict cross-sectional shapes of plasma density and pressure on the NSTX-U fusion reactor. The network runs orders of magnitude faster than existing physics-based code.


Collection of useful utility functions that I have accumulated while working on various quantum computing projects. Includes flexible state/process tomography experiments, Hamiltonian builders, many quantum logic gates, and more miscellaneous reuseable code.
[project]   [live web game]   [code]  
A 2D platformer game where the player can slow and reverse the flow of time to solve increasingly complex puzzles. Inspired by the game Portal and the movie Tenet. A live web version is hosted on this site. Made with Unity and C#.
[project]   [code]  
A board game consisting of hexagonal tiles, with each tile corresponding to a qubit. Players take turns applying quantum operations to the board to try to steer the collective board state to their target state. Graphical edition coming soon!
Cosmic string loops
In the summer of 2019, I worked with Ken Olum at the Tufts Institute of Cosmology to study the properties of smooth cosmic string loops. I began the summer running computational simulations of cosmic strings and later transitioned to working on a mathematical proof that smooth cosmic string loops will always decay.


Unresolved Research Ideas
A collection of research ideas I have tried and have not had success with - because science should encourage dissemination of negative results more!
Last Minute
[concert video (YouTube)]  
In my junior year of undergrad, some friends and I created a makeshift band to fill an open spot in CMU's annual Spring Carnival. Because we formed the band two weeks before the show date, we chose the fitting name "Last Minute". I'm all the way on stage left playing rhythm guitar! Check out 42:20 for some improvised solos at the end.