Sandia scores the best quantum hardware 137,000 times short of breaking RSA-2048
The Sandia-led QUOPS benchmark rates a quantum computer by the largest useful circuit it can finish and how fast it finishes it. Across Google, IBM and Quantinuum machines the best score was 1,824, against a chemistry target of 250 million.
The Scientist · Science desk
What happened
- Sandia National Laboratories led a team that built QUOPS, a benchmark that measures the largest computationally relevant circuit a quantum computer can successfully run and the speed at which it completes those operations.
- Quantinuum's Helios-1 trapped-ion processor ran the largest calculations, scoring more than 1,500 QUOPS on the strength of a layout in which any qubit can talk to any other, though moving the ions makes it slow.
- Google's and IBM's superconducting processors, whose qubits sit in rigid grids, were capped near 200 QUOPS but ran much faster, with Google's Willow handling 20 million operations per second.
- Breaking RSA 2048 or modeling complex molecules would take roughly 250 million to 340 million QUOPS, and the highest score any physical hardware reached was 1,824.
Why it matters
- What it costs cost Error correction costs capability: on the same Quantinuum hardware, the fault-tolerant configuration ran circuits more than 37 times smaller, even though researchers call that design route the only viable one.
- What it limits constraint The requirement figures on the record are for cryptanalysis and FeMoco-class chemistry. Nobody has worked out the equivalent number for a bank pricing financial risk or a shipper optimising logistics.
- Who has to decide decision Procurement conversations can move off qubit counts and onto two numbers, a size score and a rate, with trapped-ion and superconducting machines landing on opposite sides of them.
- What doesn't add up contradiction Nvidia, which collaborated on the work, treats the score as a reality check, while Mauerer says the benchmark cannot capture what quantum compute power really is, so its own users disagree about what a QUOPS number settles.
Next story · Product · 4d ago
Nvidia hands the Open Compute Project its Vera Rubin rack and compute tray designsConfirmed4 publishers
Error correction costs capability today, and the paper shows how much. Run as a fault-tolerant machine, one that catches and corrects its own errors, Helios-1 scored about 40 QUOPS, more than 37 times below what the same hardware managed with plain physical qubits [26]. Adding qubits makes a machine more error-prone, and researchers think error-correcting designs are the only viable future [12].
The distances to the useful targets are far larger. The low end of the requirement for breaking RSA-2048 or modeling a molecule like FeMoco, which is central to nitrogen fixing [11], is 250 million QUOPS [16]. The best physical-hardware result in the paper falls short of that by a factor of about 137,000 [24], and by about 186,000 against the top of the requirement range [25]. The fault-tolerant configuration is short by roughly 6.25 million [27]. Sandia's authors wrote that "Computational capability must grow by 5 orders of magnitude, motivating fault-tolerant approaches" [5]. Proctor told New Scientist that the turning point arrives when logical qubits reach higher QUOPS scores than ordinary qubits, after which the goal is matching those scores to useful algorithms [28].
The comparison is the part a buyer would use. An algorithm gets a QUOPS score too, and if the algorithm's score is larger than the machine's, the computation will not work; if the machine's score is comparable or larger, it ought to be useful for that problem [9]. Phys.org reported that no common scale for comparing different systems existed before this [7]. The score comes from randomized patterns of qubit operations that stand in for circuits used in some real scientific calculations, not from any customer's workload [13]. Requirement figures on the record cover cryptography and molecular chemistry [16]. Phys.org lists financial risk and logistics optimization among the problems these machines are being built for [6]. Neither report says what those problems would require.
The two figures do not rank the machines the same way. Helios-1's ions can each interact with any other, which is what let it run the largest circuits, and transporting those ions is what makes it slow [14]. Google and IBM fix their qubits in rigid grids. The grids restrict communication, cap the size score near 200 and make the machines fast [15].
The work is a preprint by Timothy Proctor and colleagues, posted to arXiv [17]. Nvidia collaborated on it, and Sam Stanwyck there described the result as a reality check [20]. Wolfgang Mauerer at the Technical University of Applied Sciences Regensburg said "These kinds of benchmarks are very useful at the moment" [21], and he also said "What quantum compute power really is, that is not yet fully understood, and that cannot be captured in the benchmark" [22].
What to watch
- Whether the 250 million to 340 million QUOPS requirement estimates survive peer review of the arXiv preprint.
- Whether Google, IBM and Quantinuum begin publishing QUOPS scores and rates for their own machines.
- Mauerer's extension: QUOPS scores for machines built on qubit types other than trapped ions and superconducting circuits.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence70
- Adoption25
- Hype gap+8
- Incentives48
- Confidence66
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
A team led by Sandia National Laboratories devised a universal testing standard, the Quantum Universal Operation Performance System (QUOPS), to provide an apples-to-apples performance measure across quantum computers, and applied it to Google, IBM and Quantinuum hardware.
- [2]
QUOPS is designed to measure the size of the largest computationally relevant quantum circuits a quantum computer can successfully run, as well as the speed at which it can complete those operations.
- [3]
The highest score achieved on physical hardware is 1,824 QUOPS, and current machines fall short of the requirement by a factor of approximately 100,000.
- [4]
Closing the gap will require far lower error rates, more physical qubits or more efficient quantum computer designs.
- [5]
The study authors write: "Computational capability must grow by 5 orders of magnitude, motivating fault-tolerant approaches."
- [6]
Quantum computers are being developed to tackle problems including exploring financial risk, modeling complex molecules and optimizing massive logistics networks.
- [7]
Until now there has been no way to benchmark the computational power of different quantum systems on the same scale to see how they compare.
- [8]
Timothy Proctor at Sandia National Laboratories in California and his colleagues say the QUOPS score reflects how many qubits a quantum computer has and how many computational operations it can execute.
- [9]
An algorithm can also be given a QUOPS score: if the algorithm's score is larger than the computer's, the computation will not work, and if the computer's score is comparable or larger, the machine ought to be useful for the problem the algorithm was designed for; a separate number, the QUOPS rate, captures how long the computation would take.
- [10]
None of the QUOPS scores for the Google, IBM and Quantinuum machines tested surpassed 2000, and FeMoco and cryptography computations would require scores well over 100,000 times larger.
- [11]
The FeMoco molecule plays a key role in nitrogen fixing.
- [12]
To become more computationally powerful, quantum computers must include more qubits, which makes them increasingly error-prone, and researchers think fault-tolerant approaches are the only viable future for quantum computers.
- [13]
Instead of evaluating individual qubits and the errors in individual operations, QUOPS runs randomized patterns of qubit operations that act as stand-ins for the types of circuits used in some real scientific calculations.
- [14]
Quantinuum's Helios-1 trapped-ion processor handled the largest calculations, achieving a QUOPS score of more than 1,500, helped by a flexible layout in which qubits can communicate with any other qubit; moving the ions takes time, which makes the system relatively slow.
- [15]
Superconducting processors from Google and IBM organize qubits in rigid grids that restrict their communication and limit their QUOPS scores to around 200, but the superconducting systems were much faster, with Google's Willow processor handling 20 million operations per second.
- [16]
Running calculations such as breaking RSA 2048 encryption or modeling complex molecules for chemistry requires a score of roughly 250 million to 340 million QUOPS.
- [17]
The work is described in a paper posted on the arXiv preprint server: Timothy Proctor et al, Benchmarking the computational power of quantum computers, arXiv (2026), DOI 10.48550/arxiv.2609.12146.
- [18]
The paper states: "Rigorous and well-motivated benchmarking, using QUOPS or similar tools, offers scientists and stakeholders the opportunity to track progress and verify computational capability independently."
- [19]
When Quantinuum's Helios-1 was tested configured to run as a fault-tolerant machine, one that catches and corrects its own errors, it earned a QUOPS score of about 40.
- [20]
Nvidia collaborated on the project, and Sam Stanwyck at Nvidia says the new work is a reality check, and that using the software could help researchers see how QUOPS scores change with small tweaks to quantum computer designs.
- [21]
"These kinds of benchmarks are very useful at the moment," says Wolfgang Mauerer at the Technical University of Applied Sciences Regensburg in Germany.
- [22]
"What quantum compute power really is, that is not yet fully understood, and that cannot be captured in the benchmark," says Mauerer.
- [23]
Mauerer says quantum computers with very different qubits than those tested could also be evaluated on QUOPS scores to give a better view of the overall state of the quantum computing industry, and that the QUOPS score and rate are unlikely to be the last word on what makes a quantum computer truly powerful.
- [24]
The 250 million QUOPS low-end requirement is about 137,000 times the best physical-hardware score of 1,824 QUOPS.
- [25]
The 340 million QUOPS top-end requirement is about 186,000 times the best physical-hardware score of 1,824 QUOPS.
- [26]
Helios-1's fault-tolerant score of about 40 QUOPS is more than 37 times below the same machine's physical-qubit score of more than 1,500 QUOPS.
- [27]
The 250 million QUOPS low-end requirement is about 6.25 million times Helios-1's fault-tolerant score of about 40 QUOPS.
- [28]
Proctor says the turning point for quantum computing will come when logical qubits, used for fault-tolerant computing, reach higher QUOPS scores than ordinary qubits, after which the goal will be to match their scores to useful algorithms.
Sources
2 independent publishers whose own reporting we read for this story.
- newscientist.comWe may finally have a way to rate a quantum computer’s usefulness
1 article · September 15, 2026
- phys.orgNew benchmark puts quantum computers to the test and reveals their limitations
1 article · September 17, 2026
Topics and entities
Topics
- Quantum chemistry simulation
- Quantum benchmarking
- Quantum Hardware
- Fault-Tolerant Quantum Computing
- Quantum cryptanalysis
Entities
- Sandia National Laboratories
- Quantinuum
- Nvidia
- arXiv
- Wolfgang Mauerer
- IBM
- Willow