Opportunities for Software and Compilation Innovation in an Open Quantum Stack

January 27-28, 2025 / Duke University
Convened by Fred Chong (University of Chicago) & Frank Mueller (North Carolina State University)

AGENDA

12pm - Lunch

1pm - Welcome / NQVL and QACTI Introduction, Ken Brown (Duke); s/w challenges, Fred Chong (Chicago), Frank Mueller (NCSU)

1:40pm - Challenges in the quantum control stack and opportunities for co-design, Ravi Naik (LBNL)

2:20pm - Open discussion: control stack

2:40pm - Break

3pm - Device-specific Optimizations for the Quantum Software Stack, Ben Hall (Infleqtion)

3:40pm - Open discussion: s/w stack challenges / device-spec. optimizations

4pm - Breakout groups organized along topics: control stack (2901), compiler (2908/2909), simulation (2910)

4:45pm - Reports from breakouts

5:30pm - Adjourn

Tuesday, January 28, 2025

8:30am - Arrive / coffee & light breakfast

9am - Quantum Compilers in Transition: Adapting from NISQ to FTQC, Yufei Ding (UCSD)

9:40am - Open discussion: compilers and QEC support

10am - Breakout 2: relation to QEC (2901), applications & compilers (2908/2909)

10:45am - Break

11am - Report back 

12pm - Lunch

1pm - Prioritization of topics and identification of risks

2pm - Writing session for initial town hall report (divided by most promising topic)

3pm - Discussion of “requirements” needed for other parts

3:30pm - Finalize town hall report writing team and identify communities that were missing at the town hall

4pm - Adjourn

SPEAKER BIOS & ABSTRACTS

Challenges in the quantum control stack and opportunities for co-design

As the performance and scale of quantum processors improves, the challenges emerge in the control and measurement infrastructure necessary to execute algorithms and protocols on these devices. Here, I outline some of the challenges facing the quantum control stack and how they can transcend traditional abstraction layers. To overcome these key obstacles, I present opportunities for co-design approaches across the stack for scientific and technical advances. These include advanced compilation techniques to take advantage of unique processor capabilities and on-the-fly randomized compiling using the open-source control platform, QubiC. Critically, these opportunities were enabled by an integrated workflow that allows for close collaboration between researchers with varied expertise.

Ravi Naik is a Research Scientist at Lawrence Berkeley National Laboratory, and Head of Measurement at the Advanced Quantum Testbed. His current research efforts include the implementation, characterization, and optimization of novel controls on superconducting qubit and qutrit processors, as well as studying the effects of noise and error on the compilation and execution of quantum algorithms and simulation. He received his Ph.D. at the University of Chicago for his research on multimode circuit quantum electrodynamics.

Device-specific Optimizations for the Quantum Software Stack

With several promising candidates for quantum advantage emerging among quantum hardware, compilation must take device-specific information into consideration in order to optimize for advantage. In this talk, we review the quantum stack and discuss examples of device-specific optimization based on various device properties, which are executed using Infleqtion's compilation software, Superstaq. First, we discuss optimized compilation based on native gate set for QSCOUT's trapped ion device that reaches down into its "hidden" gates, as well as show a technique called "SWAP mirroring" that takes advantage of the devices all-to-all connectivity. Next, we walk through a star-to-line routing algorithm that maps "stars" in quantum circuits to the "lines" common in superconducting devices. Finally, we discuss a space-time tradeoff algorithm and its advantages for large-qubit-number yet slow-gate-time neutral atom devices which takes advantage of its mid-circuit measurement capabilities.

Benjamin Hall is a senior quantum software engineer at Infleqtion where he specializes in quantum algorithms and compilation. He holds a dual Ph.D. in Physics and CMSE (Computational Math, Science, and Engineering) from Michigan State University and B.S.’s in both Physics and Applied Mathematics from the University of Rochester. His doctoral research was in the application and optimization of near-term quantum algorithms to many-body nuclear physics. His quantum background was further developed through research conducted at the NASA Ames research center, as well as Los Alamos, Oak Ridge, and Argonne National Laboratories. In his current role at Infleqtion, Ben has helped expand the capabilities of Q-CHOP, a novel quantum annealing algorithm for constrained optimization problems, invented an efficient way to detect and map star topologies, and led a project to solve sphere packing on a quantum computer to model protein chromatography. 

Quantum Compilers in Transition: Adapting from NISQ to FTQC

QUESTIONS for BREAKOUT SESSIONS 

What kind of software mechanisms can the NQVL provide that is hard to get access to elsewhere? E.g., experiment with queue scheduling and resource allocation; experiment with gate pulses; experiment with calibration routines

What kind of software mechanisms can the NQVL provide that is hard to get access to elsewhere?

What are the software elements needed to support bosonic systems?

Are there other non-standard systems that should be considered? E.g., is analog computing of interest for ion traps?

What are the new software stack elements imposed by zones and by multiple traps within a device?

Are there specific requirements for the software stack imposed by novel application scenarios? E.g., quantum-classical co-simulation

PARTICIPANT LIST 

From the QACTI team