The 5th Annual Workshop ON Control Systems for Quantum Computing (CS4Q)

Abstract: Workshop Chair, Dr. Kasra Nowrouzi, CTO of QuNorth, sets the stage for the 5th iteration of CS4Q, motivating the QEC theme for control systems in 2026, whether in isolation, in the full-stack context, or in co-design with cryogenics, etc.

Bio: Dr. Kasra Nowrouzi is the CTO of QuNorth, an independent, private, joint venture public-private partnership backed by the Novo Nordisk Foundation and the Export Investment Fund of Denmark. Based in Copenhagen, QuNorth deploys the world's most powerful error-corrected quantum computers for collaborative research with the Nordic community to accelerate quantum algorithms and technologies, and to build the quantum ecosystem.

Abstract: Exchange-only spin qubits are a promising platform for scalable quantum computing: they boast universal control from low-power baseband voltage pulses and compatibility with 300mm semiconductor fabrication. HRL's TN3 quantum processing unit (QPU) recently demonstrated classical and quantum codes with up to seven exchange-only qubits (21 spins) [1]. Alongside these results, I will also present cross-entropy benchmarking (XEB) and mirror randomized benchmarking (MRB) results on a 48-spin TN3 device in an "S=0" operation mode, which is specially designed for precise characterization of exchange-only QPUs [2]. All experiments on TN3 are stored in the digital memory of a 4K cryogenic CMOS chip, which generates analog voltage pulses and sends them directly to the mK qubit chip through a superconducting ribbon cable. Benchmarking results suggest that the majority of quantum errors in TN3 come not from microscopic noise near the qubits, but rather from non-idealities in the 4K cryo-controller and signal lines.

Bio: Stephen Carr joined the device and quantum modeling teams at HRL's Material and Microsystems Laboratory in 2023. He received a BA from Columbia University in 2015, and a PhD in theoretical physics from Harvard University in 2020. His thesis focused on multiscale ab-initio modeling of 2D materials and moiré interfaces in the Kaxiras lab. Stephen's research interests span quantum computing, condensed matter physics, and computer science.

Abstract: Building scalable, fault-tolerant quantum computers requires co-design across quantum hardware, classical control, and software. In this talk, we present an architecture for fast, digitally controlled, and scalable logical qubits based on superconducting devices. At the millikelvin stage, we integrate superconducting digital logic with on-chip pulse generation to control dual-rail superconducting qubits, enabling short syndrome rounds and substantially reducing cryogenic interconnects through digital signaling. Classical control and error decoding for a surface code are handled by a real-time decoder subsystem that achieves sub-cycle decode latencies at practical code distances and supports a universal logical gate set, including state injection and lattice surgery. We will describe the architecture and report on our progress toward implementing and experimentally demonstrating the key components. Taken together, these advances are expected to enable large-scale quantum systems with tightly coordinated classical control and error correction.

Bio: Dustin Johnson is a Consulting Systems Architect at Northrop Grumman, where they lead systems architecture and design of quantum computing systems. They received their Bachelor of Science degree in computer engineering from South Dakota School of Mines & Technology. With over 20 years of experience in systems engineering and microelectronics development for aerospace and defense, their work has focused on real-time processing systems and architectures. For the last 8 years, they have been focused on developing quantum computing systems.

Abstract: Fault-tolerant quantum computing places demands on control systems that span far beyond pulse-level sequencing: decoders must consume syndrome streams in real time, logical operations must be conditioned on decoder outcomes, and the classical resources doing this work increasingly span heterogeneous hardware from FPGA/ASIC controllers to GPUs. I will discuss real-time orchestration between the logical runtime, control system, and massively parallel decoding systems. I will close with where we see the key open interface questions between control system developers and the wider FTQC software stack.

Bio: Justin Gage Lietz is a research scientist and quantum computing architect at NVIDIA, focused on quantum error correction and integrating quantum computers with classical HPC systems. Prior to NVIDIA, Justin was on the research staff at the National Center for Computational Sciences at Oak Ridge National Laboratory, where he developed scientific applications for leadership-class supercomputers. Justin received his Ph.D. in computational nuclear physics in 2019 from Michigan State University, where his research applied high-performance computing to quantum many-body systems

Session I presenters sit together on a panel, discussing the intersection of QEC and control systems.

Abstract: This talk covers control system specific aspects of Atom Computing's recent work with the Toric

error correcting code. Functions including syndrome extraction, decoding, and corrective

feedback correspond to specific components in the control architecture. Common control flow

patterns specific to the toric QEC code will be presented. We then discuss pending

improvements: offloading control computation onto accelerators and adopting low-latency signal

distribution across the controls system. These changes are designed to give the control stack

the flexibility to accommodate rapidly evolving QEC algorithms as decoding strategies and

correction schemes continue to advance, supporting Atom Computing's full-stack approach to

neutral atom quantum computer software.

Bio: Michael Sorensen is VP of Software Engineering and Control Systems at Atom Computing,

where he leads the development of the classical compute that powers the company's neutral

atom quantum computers — from low-level control systems and hardware integration through to

the service layer APIs. Before Atom Computing, Michael held engineering and technical

leadership roles at Uber, Blue Origin, and Amazon, bringing a background in large-scale and

highly reliable systems to the challenges of building full-stack quantum hardware. He holds a

MS in Computer Science from Johns Hopkins University.

Abstract: An overview of some of the challenges of building extremely-large scale quantum control systems, with thousands of qubits and couplers, as need to construct utility-scale error corrected Quantum Computers. Discusses some of the control system challenges and IBM's approach to tackling them in upcoming error-corrected systems.

Bio: Jonathan Kaus is a Quantum Firmware Architect with more than 20 years of experience at IBM. His career spans Enterprise Systems Firmware for IBM Power systems and a broad range of advanced technology initiatives, bringing extensive expertise in firmware engineering, along with a deep understanding of enterprise computing systems and their underlying architectures.

Abstract Pending

Bio Pending

Abstract Pending

Bio Pending

Session II presenters sit together on a panel, discussing control system architecture within the context of a scalable full-stack quantum computer.

Abstract: Calibration and control orchestration for QPUs scaling to hundreds of qubits requires finer-grain control over qubit control and readout protocols to reach performance thresholds for quantum advantage. At Q-CTRL, we present an LLVM/MLIR compiler toolchain that provides a hybrid classical-quantum programming and execution model with low-level pulse control semantics, focusing on logical signalling channels.

Programs are expressed as timed operations over these channels, from which the compiler resolves a real-time execution schedule and outlines backend-agnostic device regions into concrete hardware kernels under a latency and memory-aware cost model. These kernels are dispatched from the classical host through a lightweight runtime ABI, while surrounding classical code lowers to standard LLVM. This yields portable, end-to-end optimization across heterogeneous control hardware and QPU architectures, spanning fractional-gate rewrites down to pulse-level synthesis.

By co-scheduling classical control flow with pulse-level operations and co-locating latency-critical syndrome-extraction kernels with control hardware, the toolchain provides a compilation substrate for real-time decoders and logical-qubit primitives as QEC matures. We describe our pilot implementation, an end-to-end demonstration on superconducting hardware, and the roadmap toward standardized frontends and QEC-ready multi-backend control.

Bio: Pending

Abstract: Pending

Bio: Pending

Abstract: Recent years have shown multiple demonstrations of quantum error correction (QEC) being a feasible technology. Now the challenge is to build a full stack that sustains this performance, both within a single sequence containing multiple QEC cycles and across several layers of abstraction, maintaining high up-times as quantum computers scale. From the control system perspective, Zurich Instruments has taken these requirements and began implementing solutions alongside collaborators in both academia and industry. This talk will cover some of these key collaborations and developments, including hardware developments to ensure QPUs are operating at peak performance and a scalable RoCE network protocol for tight integration with classical compute resources. In addition to these hardware focused developments, a scalable software framework with multiple layers of abstraction is needed to efficiently tune-up individual gate operations and compile quantum circuits from high-level user algorithms.

Bio: Edward Kluender is the Lead Application Scientist for quantum technologies at Zurich Instruments USA. He works closely with collaborators and partners throughout North America, ensuring the companies current and future quantum solutions are designed to keep up with and lead the field’s fast moving control requirements. Prior to joining Zurich Instruments, he received his PhD in Materials Science and Engineering at Northwestern University, focusing on nanofabrication and high-throughput combinatorial screening for a variety of applications.

Abstract: Scaling quantum processors requires a fundamental shift in control architecture. As qubit counts grow, high-latency systems cannot handle the sub-microsecond feedback needed to calibrate and correct large-scale hardware. Reaching fault tolerance depends on the tight, ultra-low-latency integration of quantum control with classical compute. This architecture powers two vital capabilities: massive speed-up and parallelization of qubit tune-up—slashing calibration cycles from days to seconds—and real-time quantum error correction using fast decoding within the coherence window. Qblox delivers these in an open, multi-resource framework (GPU, CPU, FPGA) that leaves its users with full ownership over core IP on their route to quantum advantage.

Bio: Dr. Niels Bultink is the CEO and co-founder of Qblox, leading its transformation from a TU Delft spin-off into a 200-person global quantum powerhouse. Under his leadership, Qblox supplies the scalable control stacks essential for practical quantum computing.

A pioneer in quantum information processing, Niels achieved the first-ever feedback on solid-state qubits in 2012. His technical foundation stems from more than a decade of research at QuTech on fault-tolerant quantum computing.

Niels combines technical excellence with a partner-first leadership philosophy. By fostering a grounded culture centered on collaboration and bridging deep-tech research with industrial scalability, he ensures quantum integrators worldwide can deploy at scale.

Session III presenters sit on a panel to discuss the role of control systems as instruments within the broader context of quantum computing. Control interfaces, including for QEC, will be a particular topic of focus, setting the stage for the QEC panel discussion on Thursday, September 17th, at 13:00.