Quantum Volume Testing: Setting the Steady Pace to Higher Performing Devices
When it comes to completing the statistical tests and other steps necessary for calculating quantum volume, few people have as much as experience as Dr. Charlie Baldwin.
Baldwin, a lead physicist at ĢƵ, and his team have performed the tests numerous times on three different H-Series quantum computers, which have set six industry records for measured quantum volume since 2020.
Quantum volume is a benchmark developed by IBM in 2019 to measure the overall performance of a quantum computer regardless of the hardware technology. (ĢƵ builds trapped ion systems).
Baldwin’s experience with quantum volume prompted him to share what he’s learned and suggest ways to improve the benchmark in a peer-reviewed paper published this week in .
“We’ve learned a lot by running these tests and believe there are ways to make quantum volume an even stronger benchmark,” Baldwin said.
We sat down with Baldwin to discuss quantum volume, the paper, and the team’s findings.
How is quantum volume measured? What tests do you run?
Quantum volume is measured by running many randomly constructed circuits on a quantum computer and comparing the outputs to a classical simulation. The circuits are chosen to require random gates and random connectivity to not favor any one architecture. We follow the construction proposed by IBM to build the circuits.
What does quantum volume measure? Why is it important?
In some sense, quantum volume only measures your ability to run the specific set of random quantum volume circuits. That probably doesn’t sound very useful if you have some other application in mind for a quantum computer, but quantum volume is sensitive to many aspects that we believe are key to building more powerful devices.
Quantum computers are often built from the ground up. Different parts—for example, single- and two-qubit gates—have been developed independently over decades of academic research. When these parts are put together in a large quantum circuit, there’re often other errors that creep in and can degrade the overall performance. That’s what makes full-system tests like quantum volume so important; they’re sensitive to these errors.
Increasing quantum volume requires adding more qubits while simultaneously decreasing errors. Our quantum volume results demonstrate all the amazing progress ĢƵ has made at upgrading our trapped-ion systems to include more qubits and identifying and mitigating errors so that users can expect high-fidelity performance on many other algorithms.
You’ve been running quantum volume tests since 2020. What is your biggest takeaway?
I think there’re a couple of things I’ve learned. First, quantum volume isn’t an easy test to run on current machines. While it doesn’t necessarily require a lot of qubits, it does have fairly demanding error requirements. That’s also clear when comparing progress in quantum volume tests across different platforms, .
Second, I’m always impressed by the continuous and sustained performance progress that our hardware team achieves. And that the progress is actually measurable by using the quantum volume benchmark.
The hardware team has been able to push down many different error sources in the last year while also running customer jobs. This is proven by the quantum volume measurement. For example, H1-2 launched in Fall 2021 with QV=128. But since then, the team has implemented many performance upgrades, recently achieving QV=4096 in about 8 months while also running commercial jobs.
What are the key findings from your paper?
The paper is about four small findings that when put together, we believe, give a clearer view of the quantum volume test.
First, we explored how compiling the quantum volume circuits scales with qubit number and, also proposed using arbitrary angle gates to improve performance—an optimization that many companies are currently exploring.
Second, we studied how quantum volume circuits behave without errors to better relate circuit results to ideal performance.
Third, we ran many numerical simulations to see how the quantum volume test behaved with errors and constructed a method to efficiently estimate performance in larger future systems.
Finally, and I think most importantly, we explored what it takes to meet the quantum volume threshold and what passing it implies about the ability of the quantum computer, especially compared to the requirements for quantum error correction.
What does it take to “pass” the quantum volume threshold?
Passing the threshold for quantum volume is defined by the results of a statistical test on the output of the circuits called the heavy output test. The result of the heavy output test—called the heavy output probability or HOP—must have an uncertainty bar that clears a threshold (2/3).
Originally, IBM constructed a method to estimate that uncertainty based on some assumptions about the distribution and number of samples. They acknowledged that this construction was likely too conservative, meaning it made much larger uncertainty estimates than necessary.
We were able to verify this with simulations and proposed a different method that constructed much tighter uncertainty estimates. We’ve verified the method with numerical simulations. The method allows us to run the test with many fewer circuits while still having the same confidence in the returned estimate.
How do you think the quantum volume test can be improved?
Quantum volume has been criticized for a variety of reasons, but I think there’s still a lot to like about the test. Unlike some other full-system tests, quantum volume has a well-defined procedure, requires challenging circuits, and sets reasonable fidelity requirements.
However, it still has some room for improvement. As machines start to scale up, runtime will become an important dimension to probe. IBM has proposed a metric for measuring run time of quantum volume tests (CLOPS). We also agree that the duration of the computation is important but that there should also be tests that balance run time with fidelity, sometimes called ‘time-to-solution.”
Another aspect that could be improved is filling the gap between when quantum volume is no longer feasible to run—at around 30 qubits—and larger machines. There’s recent work in this area that will be interesting to compare to quantum volume tests.
You presented these findings to IBM researchers who first proposed the benchmark. How was that experience?
It was great to talk to the experts at IBM. They have so much knowledge and experience on running and testing quantum computers. I’ve learned a lot from their previous work and publications.
There is a lot of debate about quantum volume and how long it will be a useful benchmark. What are your thoughts?
The current iteration of quantum volume definitely has an expiration date. It’s limited by our ability to classically simulate the system, so being unable to run quantum volume actually is a goal for quantum computing development. Similarly, quantum volume is a good measuring stick for early development.
Building a large-scale quantum computer is an incredibly challenging task. Like any large project, you break the task up into milestones that you can reach in a reasonable amount of time.
It's like if you want to run a marathon. You wouldn’t start your training by trying to run a marathon on Day 1. You’d build up the distance you run every day at a steady pace. The quantum volume test has been setting our pace of development to steadily reach our goal of building ever higher performing devices.
About ĢƵ
ĢƵ, the world’s largest integrated quantum company, pioneers powerful quantum computers and advanced software solutions. ĢƵ’s technology drives breakthroughs in materials discovery, cybersecurity, and next-gen quantum AI. With over 500 employees, including 370+ scientists and engineers, ĢƵ leads the quantum computing revolution across continents.
ĢƵ and GE Vernova’s Advanced Research Center are launching a collaboration to apply quantum computing to one of the research community’s most consequential challenges: understanding the behavior behind high-temperature superconductivity. Advances in this field could, over the long term, transform the energy sector by enabling lighter generators, more compact and powerful motors, and electricity infrastructure that carries more power and better withstands faults.
The research will focus on Fermi–Hubbard modeling, a framework for studying how interacting electrons give rise to complex material properties. By combining ĢƵ’s full-stack quantum computing capabilities—including access to the company’s Helios computer—with GE Vernova’s expertise in material science and engineering simulation applied to industrial applications, the collaboration will bring an industry perspective to a promising area of quantum research.
Why superconductors matter for energy
As energy systems grow, the size, weight, and performance of electrical equipment become increasingly important constraints. High-temperature superconductors offer potential ways to address those limits through their ability to carry large electrical currents without resistance under suitable operating conditions.
Potential applications span power generation, electricity grids, and industrial electrification. Superconducting generators could reduce the weight and footprint of industrial equipment. Superconducting windings could also help make utility generators and industrial or marine propulsion motors more compact and efficient. In electricity networks, superconducting cables could carry more power through constrained spaces, while fault current limiters could help protect equipment during short circuits.
Realizing those benefits more widely depends on overcoming challenges in cooling, material performance, manufacturing, and cost. Better understanding of the physics behind high-temperature superconductivity could help inform the search for materials with more practical operating characteristics.
Why quantum computing may offer new tools for understanding high-temperature superconductivity
High-temperature superconductivity involves the collective behavior of strongly interacting electrons. Accurately modeling these systems is exceptionally difficult: the interactions that make them scientifically interesting also make important aspects of their behavior computationally challenging.
The Fermi–Hubbard model captures essential features of those interactions in a simplified mathematical framework. Although it is not a complete description of a material used in industrial applications, it provides a foundation for investigating mechanisms associated with high-temperature superconductivity.
Quantum computing offers a promising way to explore the physics. ĢƵ researchers have already measured superconducting pairing correlations—including s-wave and d-wave pairing—in Fermi–Hubbard models simulated on the Helios quantum computer. These results demonstrate the ability to prepare and probe quantum states relevant to superconductivity research, establishing a foundation for further investigation.
The collaboration with GE Vernova will build on that foundation, bringing industrial expertise to the selection and interpretation of research questions.
Connecting quantum capabilities with industrial expertise
The research will be enabled through access to ĢƵ’s full-stack capabilities, including cloud-based access to Helios, the world’s most accurate quantum computer1, and Nexus, ĢƵ’s developer platform. Nexus connects quantum hardware with the software tools researchers need to develop and run applications in a unified environment.
GE Vernova’s Advanced Research Center brings the expertise in material science, engineering, and industrial applications to the research and development activity.
“High-temperature superconductors could help us overcome important limits on the size, weight, and efficiency of electrical equipment, but understanding and improving these materials remains a significant scientific challenge,” said Jonathan Owens, PhD, Senior Scientist, Computational Materials Physics at GE Vernova. “By combining our materials simulation and industrial expertise with ĢƵ’s quantum computing capabilities, we can explore new approaches to these difficult problems while building the expertise to assess where quantum computing can make a meaningful contribution.”
The collaboration also extends GE Vernova’s exploration of quantum computing into a new area of materials research while supporting the development of its internal quantum talent. Its previous includes investigations of quantum algorithms for carbon-capture materials using simulated quantum processors.
Building toward industrial value
Translating insights from simplified models into practical materials requires sustained scientific and engineering work. Industrial expertise helps connect that effort to the properties, operating conditions, and manufacturing considerations that ultimately determine a material’s usefulness.
For ĢƵ, this collaboration is an opportunity to advance quantum simulation with an industrial collaborator that understands both the potential applications and the barriers to adoption. It brings together demonstrated computing capabilities, a challenging materials problem, and the expertise needed to connect scientific progress with industrial requirements.
The research collaboration aims to deepen understanding of this field and explore how quantum computing may contribute to future energy technologies.
1 Based on two-qubit gate fidelity as of December 31, 2025.
From Oracle's Quantum AI Compute beta program to upcoming Helios installation milestones and Oracle AI World, our partnership continues to gain momentum.
When ĢƵ and Oracle announced a strategic partnership earlier this year, the goal was not simply to deploy a quantum computer in the cloud. It was to create a practical path for bringing quantum computing into the enterprise environments where AI, data, and high-performance computing (HPC) already come together.
By combining Oracle Cloud Infrastructure (OCI)'s AI, data, and cloud capabilities with Helios, ĢƵ's most advanced quantum computer, we're helping lay the foundation for hybrid quantum AI computing at enterprise scale. The partnership reflects our belief that some of the most valuable future applications will emerge when quantum computing, AI and classical HPC work together to tackle problems that remain difficult to solve today.
Oracle Opens Early Access for Quantum AI Compute
That vision is now being put into practice. This month, Oracle launched the , giving select organizations an opportunity to begin evaluating how quantum computing fits into existing workflows and computational challenges.
Through the program, approved participants will gain cloud-based access to Helios, the world's most accurate commercial quantum computer, alongside OCI's GPU and HPC infrastructure. For organizations exploring quantum computing, this creates a practical starting point: access to industry-leading hardware within an enterprise cloud environment, without the need to build specialized infrastructure of their own.
The goal is to help organizations understand where hybrid quantum AI workflows may create value across areas such as drug discovery, materials science, energy, financial modeling, and optimization.
For many organizations, the starting point is not quantum technology itself. It's a business or research problem. Where are today's approaches reaching meaningful limits? Which workloads consume disproportionate resources? And where could a new computational capability create an advantage?
Organizations interested in participating can apply through Oracle's OCI Preview Programs . Oracle will review applications and select participants based on eligibility, capacity and program requirements.
ĢƵ and Oracle Leadership Takes the Stage at Oracle AI World
The opportunity for hybrid quantum-AI computing will be a major topic at .
In a main stage session, "," Larry Stack, Chief Commercial Officer at ĢƵ, will join Mahesh Thiagarajan, Executive Vice President and Head of Oracle Cloud Infrastructure, to discuss how quantum computing, AI and classical HPC can work together to address some of the world's most computationally demanding challenges.
The discussion will focus on a core question many organizations are asking: where should they start? Rather than leading with technology, the conversation will explore how organizations can identify high-value problems, establish today's performance baseline, and evaluate where quantum computing ultimately provides an advantage.
Later in the day, ĢƵ and Oracle experts will at how bringing Helios into OCI provides managed access to quantum computing natively within an enterprise cloud environment to support hybrid quantum-AI workflows.
Building Momentum Toward Hybrid Quantum-AI Computing
The collaboration between ĢƵ and Oracle continues to progress from vision to execution.
With Helios advancing toward installation, Oracle's Quantum AI Compute program now open for early access, and enterprise interest growing around hybrid computing, the pieces are coming together. Organizations now have a practical way to begin exploring quantum computing through cloud-based access to Helios alongside AI, HPC, and existing enterprise workflows.
We look forward to continuing the conversation at Oracle AI World 2026 and working with customers who want to better understand where hybrid quantum-AI computing can create real-world value.
We're excited to continue that conversation at Oracle AI World 2026 and to help customers begin exploring what's possible through the OCI Quantum AI Compute Beta Program.
By: David Hayes, Sr. Director of Computational Design and Integrated Software
Two recent announcements from ĢƵ and Google mark important progress toward fault-tolerant quantum computing.
Because both announcements focus on reducing errors, it may be tempting to compare the numbers directly. But these results are not the same. They measure different parts of the fault-tolerance challenge and answer fundamentally different technical questions.
This week, we a world-record error-correction result, using our error-correction architecture on Helios to deliver a logical fidelity of 99.9995%, equivalent to just five errors per million operations and representing a 100x reduction in errors compared to the best published result across any other modality—reported by .
Last week, Google a 99.99% magic-state fidelity result using a new cultivation technique, citing a 40x reduction in errors.
As fault-tolerant quantum computing matures, understanding what is being measured becomes just as important as understanding the headline number itself. Google's work measured the fidelity of a resource used in certain fault-tolerant operations. Our work measured the fidelity of the logical computation that consumes those resources.
Across the published results, ĢƵ demonstrates state-of-the-art performance in fault-tolerance, combining record logical computational fidelity with substantial advantages in resource efficiency. Understanding starts with examining what each study measured, what it achieved, and the resources it required.
ĢƵ's World-Record Result: Measuring the Accuracy of a Logical Computation
This week’s result measured how accurately an error-corrected logical quantum system can execute a computation.
To achieve the result, we used a well-known “color code” within our error-correction architecture. The error rate is 10 times lower than ĢƵ’s own world record achieved last month and 100 times lower than the best published result across all other hardware modalities—the 2024 Google result.
The study also demonstrated that logical errors fell exponentially as code distance increased. Code distance is a key measure of how quantum information can be protected as an error-correction system scales.
Each increase tested in the study reduced the remaining logical error rate by a record factor of 12, demonstrating exponentially effective error suppression as the code grows, a critical capability for scalable fault-tolerant quantum computing.
Together, these results demonstrate how ĢƵ's highly accurate trapped-ion hardware and flexible logical architecture can drive rapid gains in computational accuracy. They also show that fault-tolerant performance can be achieved with substantially fewer physical qubits than competing modalities.
The result places ĢƵ within the target logical-performance range required for Apollo, our planned universal fault-tolerant quantum computer targeted for 2029.
Google's Result: Advancing Magic-State Preparation
Google's latest paper focused on a different aspect of fault tolerance: the preparation of high-fidelity magic states.
Magic states are an important resource required for certain operations in a fault-tolerant quantum computer. Generating them efficiently is widely recognized as one of the key engineering challenges facing large-scale fault-tolerant systems, particularly for certain modalities.
Google's work introduced a new approach known as cultivation, designed to improve the fidelity of magic-state preparation while reducing the resources required to produce those states. The study demonstrated a magic-state fidelity of 99.99% using a protocol that retained approximately 8% of attempts while discarding the remainder. This post-selection process was a key contributor to the reported 40x reduction in error rate.
The study did not measure the fidelity of a larger logical computation that incorporates those magic states into an end-to-end algorithm. Instead, it focused specifically on the creation and refinement of the resource itself.
That distinction is important because preparing a high-fidelity resource and demonstrating the performance of a logical computation are different milestones on the path to fault tolerance.
Directly Comparing ĢƵ and Google’s Magic-State Results
ĢƵ's latest logical computation result should not be compared directly with Google's latest magic-state result. It should be compared against the distance scaling result published by Google in 2024.
That said, we have published studies measuring an analogous magic-state capability examined in Google’s latest work.
, we achieved 99.993% magic-state fidelity with an 85% retention rate, compared with the 99.99% fidelity and approximately 8% retention reported in Google’s latest work. Our result also required fewer physical qubits and fewer operational steps, resulting in significantly lower space-time overhead.
That combination of smaller qubit and operational steps footprint, along with substantially higher retention means fewer resources are consumed preparing each usable magic state.
As fault-tolerant systems scale, those efficiency advantages become increasingly important because they directly affect the amount of hardware required to perform useful computations.
The Bigger Picture
Taken together, these results establish ĢƵ’s leadership in demonstrated fault-tolerant performance: record logical computation fidelity, alongside a directly comparable magic-state result that achieves higher fidelity with substantially lower resource overhead.
That combination—computational accuracy and the efficiency with which it is achieved—is critical to delivering reliable answers to useful problems at a practical scale.
Ultimately, customers do not buy two-qubit gate fidelities or magic states. They buy answers.
Forward-Looking Statements
This blog post contains forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995. These include statements about our technology roadmap, our expected logical performance, and the planned timing and capabilities of future systems, including Apollo. Words such as "target," "expect," "plan," and similar expressions identify forward-looking statements. These statements are based on our current expectations and involve risks and uncertainties that could cause actual results to differ materially. Those risks include challenges in developing, manufacturing, and scaling our quantum computing systems. For a discussion of these and other risks, see the "Risk Factors" section of our most recent Quarterly Report on Form 10-Q and our other filings with the SEC, available at ir.quantinuum.com. Forward-looking statements speak only as of the date of this post, and we undertake no obligation to update them except as required by law.