Quantum Computers Will Make AI Better
Quantum computers will drive AI to new heights, enabling better accuracy and therefore better performance, and scalable sustainable growth.
Today’s LLMs are often impressive by past standards – but they are far from perfect
Quietly, and determinedly since 2019, we’ve been working on Generative Quantum AI. Our early focus on building natively quantum systems for machine learning has benefitted from and been accelerated by access to the world’s most powerful quantum computers, and quantum computers that cannot be classically simulated.
Our work additionally benefits from being very close to our Helios generation quantum computer, built in Colorado, USA. Helios is 1 trillion times more powerful than our H2 System, which is already significantly more advanced than all other quantum computers available.
While tools like ChatGPT have already made a profound impact on society, a critical limitation to their broader industrial and enterprise use has become clear. Classical large language models (LLMs) are computational behemoths, prohibitively huge and expensive to train, and prone to errors that damage their credibility.
Training models like ChatGPT requires processing vast datasets with billions, even trillions, of parameters. This demands immense computational power, often spread across thousands of GPUs or specialized hardware accelerators. The environmental cost is staggering—simply training GPT-3, for instance, consumed nearly 1,300 megawatt-hours of electricity, equivalent to the annual energy use of 130 average U.S. homes.
This doesn’t account for the ongoing operational costs of running these models, which remain high with every query.
Despite these challenges, the push to develop ever-larger models shows no signs of slowing down.
Enter quantum computing. Quantum technology offers a more sustainable, efficient, and high-performance solution—one that will fundamentally reshape AI, dramatically lowering costs and increasing scalability, while overcoming the limitations of today's classical systems.
Quantum Natural Language Processing: A New Frontier
At Quantinuum we have been maniacally focused on “rebuilding” machine learning (ML) techniques for Natural Language Processing (NLP) using quantum computers.
Our research team has worked on translating key innovations in natural language processing — such as word embeddings, recurrent neural networks, and transformers — into the quantum realm. The ultimate goal is not merely to port existing classical techniques onto quantum computers but to reimagine these methods in ways that take full advantage of the unique features of quantum computers.
We have a deep bench working on this. Our Head of AI, Dr. Steve Clark, previously spent 14 years as a faculty member at Oxford and Cambridge, and over 4 years as a Senior Staff Research Scientist at DeepMind in London. He works closely with Dr. Konstantinos Meichanetzidis, who is our Head of Scientific Product Development and who has been working for years at the intersection of quantum many-body physics, quantum computing, theoretical computer science, and artificial intelligence.
A critical element of the team’s approach to this project is avoiding the temptation to simply “copy-paste”, i.e. taking the math from a classical version and directly implementing that on a quantum computer.
This is motivated by the fact that quantum systems are fundamentally different from classical systems: their ability to leverage quantum phenomena like entanglement and interference ultimately changes the rules of computation. By ensuring these new models are properly mapped onto the quantum architecture, we are best poised to benefit from quantum computing’s unique advantages.
These advantages are not so far in the future as we once imagined – partially driven by our accelerating pace of development in hardware and quantum error correction.
Making computers “talk”- a short history
The ultimate problem of making a computer understand a human language isn’t unlike trying to learn a new language yourself – you must hear/read/speak lots of examples, memorize lots of rules and their exceptions, memorize words and their meanings, and so on. However, it’s more complicated than that when the “brain” is a computer. Computers naturally speak their native languages very well, where everything from machine code to Python has a meaningful structure and set of rules.
In contrast, “natural” (human) language is very different from the strict compliance of computer languages: things like idioms confound any sense of structure, humor and poetry play with semantics in creative ways, and the language itself is always evolving. Still, people have been considering this problem since the 1950’s (Turing’s original “test” of intelligence involves the automated interpretation and generation of natural language).
Up until the 1980s, most natural language processing systems were based on complex sets of hand-written rules. Starting in the late 1980s, however, there was a revolution in natural language processing with the introduction of machine learning algorithms for language processing.
Initial ML approaches were largely “statistical”: by analyzing large amounts of text data, one can identify patterns and probabilities. There were notable successes in translation (like translating French into English), and the birth of the web led to more innovations in learning from and handling big data.
What many consider “modern” NLP was born in the late 2000’s, when expanded compute power and larger datasets enabled practical use of neural networks. Being mathematical models, neural networks are “built” out of the tools of mathematics; specifically linear algebra and calculus.
Building a neural network, then, means finding ways to manipulate language using the tools of linear algebra and calculus. This means representing words and sentences as vectors and matrices, developing tools to manipulate them, and so on. This is precisely the path that researchers in classical NLP have been following for the past 15 years, and the path that our team is now speedrunning in the quantum case.
Quantum Word Embeddings: A Complex Twist
The first major breakthrough in neural NLP came roughly a decade ago, when vector representations of words were developed, using the frameworks known as Word2Vec and GloVe (Global Vectors for Word Representation). In a recent paper, our team, including Carys Harvey and Douglas Brown, demonstrated how to do this in quantum NLP models – with a crucial twist. Instead of embedding words as real-valued vectors (as in the classical case), the team built it to work with complex-valued vectors.
In quantum mechanics, the state of a physical system is represented by a vector residing in a complex vector space, called a Hilbert space. By embedding words as complex vectors, we are able to map language into parameterized quantum circuits, and ultimately the qubits in our processor. This is a major advance that was largely under appreciated by the AI community but which is now rapidly gaining interest.
Using complex-valued word embeddings for QNLP means that from the bottom-up we are working with something fundamentally different. This different “geometry” may provide advantage in any number of areas: natural language has a rich probabilistic and hierarchical structure that may very well benefit from the richer representation of complex numbers.
The Quantum Recurrent Neural Network (RNN)
Another breakthrough comes from the development of quantum recurrent neural networks (RNNs). RNNs are commonly used in classical NLP to handle tasks such as text classification and language modeling.
Our team, including Dr. Wenduan Xu, Douglas Brown, and Dr. Gabriel Matos, implemented a quantum version of the RNN using parameterized quantum circuits (PQCs). PQCs allow for hybrid quantum-classical computation, where quantum circuits process information and classical computers optimize the parameters controlling the quantum system.
In a recent experiment, the team used their quantum RNN to perform a standard NLP task: classifying movie reviews from Rotten Tomatoes as positive or negative. Remarkably, the quantum RNN performed as well as classical RNNs, GRUs, and LSTMs, using only four qubits. This result is notable for two reasons: it shows that quantum models can achieve competitive performance using a much smaller vector space, and it demonstrates the potential for significant energy savings in the future of AI.
In a similar experiment, our team partnered with Amgen to use PQCs for peptide classification, which is a standard task in computational biology. Working on the Quantinuum System Model H1, the joint team performed sequence classification (used in the design of therapeutic proteins), and they found competitive performance with classical baselines of a similar scale. This work was our first proof-of-concept application of near-term quantum computing to a task critical to the design of therapeutic proteins, and helped us to elucidate the route toward larger-scale applications in this and related fields, in line with our hardware development roadmap.
Quantum Transformers - The Next Big Leap
Transformers, the architecture behind models like GPT-3, have revolutionized NLP by enabling massive parallelism and state-of-the-art performance in tasks such as language modeling and translation. However, transformers are designed to take advantage of the parallelism provided by GPUs, something quantum computers do not yet do in the same way.
In response, our team, including Nikhil Khatri and Dr. Gabriel Matos, introduced “Quixer”, a quantum transformer model tailored specifically for quantum architectures.
By using quantum algorithmic primitives, Quixer is optimized for quantum hardware, making it highly qubit efficient. In a recent study, the team applied Quixer to a realistic language modeling task and achieved results competitive with classical transformer models trained on the same data.
This is an incredible milestone achievement in and of itself.
This paper also marks the first quantum machine learning model applied to language on a realistic rather than toy dataset.
This is a truly exciting advance for anyone interested in the union of quantum computing and artificial intelligence, and is in danger of being lost in the increased ‘noise’ from the quantum computing sector where organizations who are trying to raise capital will try to highlight somewhat trivial advances that are often duplicative.
Quantum Tensor Networks. A Scalable Approach
Carys Harvey and Richie Yeung from Quantinuum in the UK worked with a broader team that explored the use of quantum tensor networks for NLP. Tensor networks are mathematical structures that efficiently represent high-dimensional data, and they have found applications in everything from quantum physics to image recognition. In the context of NLP, tensor networks can be used to perform tasks like sequence classification, where the goal is to classify sequences of words or symbols based on their meaning.
The team performed experiments on our System Model H1, finding comparable performance to classical baselines. This marked the first time a scalable NLP model was run on quantum hardware – a remarkable advance.
The tree-like structure of quantum tensor models lends itself incredibly well to specific features inherent to our architecture such as mid-circuit measurement and qubit re-use, allowing us to squeeze big problems onto few qubits.
Since quantum theory is inherently described by tensor networks, this is another example of how fundamentally different quantum machine learning approaches can look – again, there is a sort of “intuitive” mapping of the tensor networks used to describe the NLP problem onto the tensor networks used to describe the operation of our quantum processors.
What we’ve learned so far
While it is still very early days, we have good indications that running AI on quantum hardware will be more energy efficient.
We recently published a result in “random circuit sampling”, a task used to compare quantum to classical computers. We beat the classical supercomputer in time to solution as well as energy use – our quantum computer cost 30,000x less energy to complete the task than Frontier, the classical supercomputer we compared against.
We may see, as our quantum AI models grow in power and size, that there is a similar scaling in energy use: it’s generally more efficient to use ~100 qubits than it is to use ~10^18 classical bits.
Another major insight so far is that quantum models tend to require significantly fewer parameters to train than their classical counterparts. In classical machine learning, particularly in large neural networks, the number of parameters can grow into the billions, leading to massive computational demands.
Quantum models, by contrast, leverage the unique properties of quantum mechanics to achieve comparable performance with a much smaller number of parameters. This could drastically reduce the energy and computational resources required to run these models.
The Path Ahead
As quantum computing hardware continues to improve, quantum AI models may increasingly complement or even replace classical systems. By leveraging quantum superposition, entanglement, and interference, these models offer the potential for significant reductions in both computational cost and energy consumption. With fewer parameters required, quantum models could make AI more sustainable, tackling one of the biggest challenges facing the industry today.
The work being done by Quantinuum reflects the start of the next chapter in AI, and one that is transformative. As quantum computing matures, its integration with AI has the potential to unlock entirely new approaches that are not only more efficient and performant but can also handle the full complexities of natural language. The fact that Quantinuum’s quantum computers are the most advanced in the world, and cannot be simulated classically, gives us a unique glimpse into a future.
The future of AI now looks very much to be quantum and Quantinuum’s Gen QAI system will usher in the era in which our work will have meaningful societal impact.
About Quantinuum
Quantinuum, the world’s largest integrated quantum company, pioneers powerful quantum computers and advanced software solutions. Quantinuum’s technology drives breakthroughs in materials discovery, cybersecurity, and next-gen quantum AI. With over 500 employees, including 370+ scientists and engineers, Quantinuum leads the quantum computing revolution across continents.
Quantinuum’s real world experiment, on the world’s most powerful quantum computer, is the largest of its kind— so large that no amount of classical computing could match it
In 1911, a student working under famed physicist Heike Kamerlingh Onnes made a discovery that would rewire our understanding of electricity. The student was studying the electrical resistance of wires, a seemingly simple question that held secrets destined to surprise the world.
Kamerlingh Onnes had recently succeeded in liquefying helium, a feat so impressive it earned him the Nobel Prize in Physics two years later. With this breakthrough, scientists could now immerse other materials in a cold bath of liquid Helium, cooling things to unprecedented temperatures and observing their behavior.
Many theories existed about what would happen to a wire at such low temperatures. Lord Kelvin predicted that electrons would freeze in place, making the resistance infinite and stopping the conduction of electricity. Others expected resistance to decrease linearly with temperature—a hypothesis that led to thermometer designs still in use today.
When the student cooled a mercury wire to 3.6 degrees above absolute zero, he found something remarkable: the electrical resistivity suddenly vanished.
Onnes quickly devised an ingenious experiment: as a diligent researcher, he knew that he needed to validate these surprising findings. He took a closed loop of wire, set a current running through it, and watched as it flowed endlessly without fading—a type of perpetual motion that seemed to defy everything we know about physics. And so, superconductivity was born.
More than a century later, all known superconductors still require extreme conditions like brutal cold or high pressure. If we could instead design a material that superconducts at room temperature, and under normal conditions, our world would be profoundly reshaped. “Room temperature superconductivity”, as it is generally called, would enable a raft of technological breakthroughs from affordable MRI machines to nearly lossless power grids.
Designing such a material means answering many open questions, and scientists are pursuing diverse strategies to find answers. One promising approach is light-induced superconductivity. In one astonishing study, researchers at the Max Planck Institute in Hamburg used light to entice a material that normally superconducts at roughly -180 °C to superconduct at room temperature - but only for a few picoseconds. This effect raised new questions: how does light achieve something that scientists have been grappling with for decades? What is the microscopic mechanism behind this phenomenon? Could understanding it unlock practical room-temperature superconductors?
Nature’s language is mathematics and mathematics is the language of the world’s most powerful quantum computer, Helios
Physics is a surprisingly profound field when you stop to think about it. At its core lies the idea that nature speaks the language of mathematics—and that by discovering the right equations, we can reveal her secrets. As bold as that sounds, history has proven it true time and again. Whenever we peek behind the veil; mathematics is there.
To understand a phenomena like superconductivity, physicists first need a mathematical model, or a set of equations that describe how it works. With the right model, they can predict and even design new superconductors that operate under more practical conditions. This is a key frontier in the search for room temperature superconductors, one of science’s holy grails.
Since the discovery of superconductivity, a lot of work has gone into finding this right model – one that can act as a sort of ‘Rosetta stone’ for harnessing this phenomenon. One of the best bets for describing high temperature superconductors like the one in the Hamburg study is called the “non-equilibrium Fermi-Hubbard” model, which describes how electrons interact and move in a crystal.
A surprising element of models that describe superconductivity is the prediction that electrons ‘pair up’ when the material becomes superconducting, dancing around in a waltz, two at a time. These pairs are referred to as “cooper pairs” after the famous physicist Leon Cooper. Now, scientists studying superconductors look for “pairing correlations”, a key signature of superconductivity.
Even armed with the Fermi-Hubbard model, light-induced superconductivity has been very difficult to study. The world’s most powerful supercomputers can only handle very small versions, limiting their utility. Even quantum platforms, like analog simulators, limit researchers to observing ‘average’ quantities and obscuring the microscopic details that are crucial for unravelling this mystery.
Light-induced superconductivity has proved challenging to study with quantum computers as well, as doing so requires low error rates, many qubits, and extreme flexibility to measure the fickle symptoms of superconductivity.
That was, until now: Quantinuum’s Helios is one of the first machines in the world able to handle the complexity of the non-equilibrium Fermi-Hibbard model at scales previously out of reach.
Hopping across the lattice and connecting the dots
Before Helios, we were limited to small explorations of this model, stalling research on this critical frontier. Now, with Helios, we have a quantum computer uniquely suited for this problem. With a novel fermionic encoding and using up to 90 qubits (72 system qubits plus 18 ancilla), Helios can simulate the dynamics of a 6×6 lattice — a system so large that its full quantum state spans over 2^72 dimensions.
Using Helios to study a system like this offers researchers a sort of “qubit-based laboratory.” Capable of handling complex quantum mechanical effects better than classical computers, Helios allows researchers to thoroughly explore phenomena like this without wasting expensive laboratory time and materials, or spending lots of money and energy running it on a supercomputer.
Our qubit-based laboratory is a dream come true for several reasons. First, it allows arbitrary state preparation – preparing states far from equilibrium, a challenging task for classical computers. Second, it allows for meaningfully long ‘dynamical simulation’ – seeing how the state evolves in time as entanglement spreads and complexity increases. This is notoriously difficult for classical computers, in part due to their difficulty with handling distinctly quantum phenomena like entanglement. Finally, it allows for flexible measurements and experimental parameters – you can measure any observable, including critical “off-diagonal” observables that carry the signature of superconductivity, and simulate any system, such as those with laser pulses or electric fields.
This last point is the most significant. While analog quantum simulators, like cold atom systems, can take snapshots of atom positions or measure densities, they struggle with off-diagonal observables—the very ones that signal the formation of Cooper pairs in superconductors.
Breaking new ground: a light-induced pairing
In our work, we've simulated three different regimes of the Fermi-Hubbard model and successfully measured non-zero superconducting pairing correlations — a first for any quantum computing platform.
We began by preparing a low-energy state of the model at half-filling — a standard benchmark for testing quantum simulations. Then, using simulated laser pulses or electric fields, we perturbed the system and observed how it responded.
After these perturbations, we measured a notable increase in the so-called “eta” pairing correlations, a mathematical signature of superconducting behavior. These results prove that our computers can help us understand light-induced superconductivity, such as the results from the Max Planck researchers. However, unlike those physical experiments, Helios offers a new level of control and insight. By tuning every aspect of the simulation — from pulse shape, to field strength, to lattice geometry — researchers can explore scenarios that are completely inaccessible to real materials or analog simulators.
Looking to a future where superconductors permeate our lives
Why does any of this matter? If we could predict which materials will become superconducting — and at what temperature, field, or current — it would transform how we search for new superconductors. Instead of trial-and-error in the lab, scientists could design and test new materials digitally first, saving huge amounts of time and money.
In the long run, Helios and its successors will become essential tools for materials science — not just confirming theories but generating new ones. And perhaps, one day, they’ll help us crack the code behind room-temperature superconductors.
Until then, the quantum revolution continues, one entangled pair at a time.
We’re pleased to introduce Helios, a technological marvel redefining the possible.
Building on its predecessor H2, which has already breached quantum advantage, Helios nearly doubles the qubit count and surpasses H2’s industry-leading fidelity, pushing further into the quantum advantage regime than any system before it. With unprecedented capability across its full stack, Helios is the most powerful quantum computer in the world.
“Helios is a true marvel—a seamless fusion of hardware and software, creating a platform for discovery unlike any other.”- Dr. Rajeeb Hazra, CEO
Helios’ groundbreaking design and advanced software stack bring quantum programming closer than ever to the ease and flexibility of classical computing—positioning Helios to accelerate commercial adoption. Even before its public debut, Helios had already demonstrated its capabilities as the world’s first enterprise-grade quantum computer. During a two-month early access program, select partners including SoftBank Corp. and JPMorgan Chase conducted commercially relevant research. We also leveraged Helios to perform large-scale simulations in high-temperature superconductivity and quantum magnetism—both with clear pathways to real-world industry applications.
Helios is now available to all customers through our cloud service and on-premise offering, including an option to integrate with NVIDIA GB200 for applications targeting specific end markets.
A Stellar Quantum Computer
“You would need to harvest every star in the universe to power a classical machine that could do the same calculations we did with Helios."
- Dr. Anthony Ransford, Helios Lead Architect
As we detailed in a benchmarking paper, Helios sets a new standard for quantum computing performance with the highest fidelity ever released to the market. It features 98 fully connected physical qubits with single-qubit gate fidelity of 99.9975% and two-qubit gate fidelity of 99.921% across all qubit pairs—making it the most accurate commercial quantum computer in the world.
Our fidelity shines in system-level benchmarks, such as Random Circuit Sampling (RCS), famously used by Google to demonstrate quantum supremacy when it performed an RCS task that would take a classical computer “10 septillion years” to replicate. Now, RCS serves as both a benchmark and the minimum standard for serious competitors in the market. Frequently missed in this conversation, however, is the importance of fidelity, or accuracy. That's why, when benchmarking Helios using RCS, we report the fidelity achieved by Helios on circuits of varying complexity (with complexity quantified by power requirements for classical simulation).
Our results show a classical supercomputer would require more power than the Sun—or, in fact, the combined power of all stars in the visible universe—to complete the same task in the same amount of time. In contrast, Helios achieved it using roughly the power of a single data center rack.
Like its predecessors, H1 and H2, Helios is designed to improve fidelity and overall system performance over time while sustaining competitive leadership through the launch of its successor.
Qubits at a Crossroads
"When I first saw the rotatable ion storage ring with a junction and gating legs sketched on a napkin, I loved the idea for its simplicity and efficiency. Seeing it finally realized after all of the team’s hard work has been truly incredible."
- Dr. John Gaebler, Fellow and Chief Scientist, Quantinuum
The Helios ion trap uses tiny currents to generate electromagnetic fields that hold single atomic ions (qubits) hovering above the trap for computation. We introduced a first-of-its-kind “junction”, which acts like a traffic intersection for qubits, enabling efficient routing and improved reliability. This is not only the first commercial implementation of this engineering triumph but it also allows our QCCD (Quantum Charged Coupled Device) architecture to scale, with future systems featuring hundreds of junctions arranged like a city street grid.
Whereas predecessor systems routed qubits using “physical swaps,” requiring sequential sorting, cooling, and gating that prevented parallel operations, the Helios QPU instead resembles a classical architecture with dedicated memory, cache, and computational zones. Like a spinning hard drive, the Helios QPU rotates qubits through ring storage (memory), passes them through the junction into the cache, moves them to logic zones for gating, and moves them to the leg storage while the next batch is processed. Sorting can now be done in parallel with cooling operations, resulting in a processor that is faster and less error prone. This parallelism will become a hallmark of Quantinuum’s future generations, enabling faster operating speeds.
Animation: This triumph of engineering demonstrates exquisite control over some of nature’s smallest particles in a way the world has never seen; one colleague likened the ions to a “little marching band.”
Quantinuum’s QCCD provides full all-to-all connectivity, giving the Helios QPU significant advantages over “fixed qubit” architectures, such as those used in superconducting systems. Its ability to physically move qubits around and entangle any qubit with any other qubit enables algorithms and error-correcting codes that are functionally impossible for fixed qubit architectures.
We made another “tiny” but significant change: we switched our qubits from ytterbium to barium. Whereas ytterbium largely relied on ultraviolet lasers that are expensive and hard on other components, barium can be manipulated with lasers in the visible part of the spectrum, where mature industrial technology exists, providing a more affordable, reliable and scalable commercial solution.
Barium also naturally allows the quantum computer to detect and remove a certain type of error, known as leakage, at the atomic level. By addressing this error directly, programmers can enhance the performance of their computation.
Delivered on Time – in Real Time
As announced earlier this year, Helios launched with a completely new stack equipped with a new software environment that makes quantum programming feel as intuitive as classical development.
Our new stack also features a real-time engine that massively improves our capability. With a real-time control system, we are evolving from static, pre-planned circuits to dynamic quantum programs that respond to results on the fly. We can now, for the first time on a quantum computer, interleave GPU-accelerated classical and quantum computations in a single program.
Our real-time engine also means we have dynamic transport – routing qubits as the moment demands reduces time to solution and diminishes the impact of memory errors.
Programmers can now use our new quantum programming language, Guppy, to write dynamic circuits that were previously impossible. By combining Guppy with our real-time engine, developers can leverage arbitrary control flow driven by quantum measurements, as well as full classical computation—including loops, higher-order functions, early exits, and dynamic qubit allocation. Far from being mere conveniences, these capabilities are essential stepping stones toward achieving fault-tolerant quantum computing at scale—putting us decisively ahead of the competition.
Fully compatible with industry standards like QIR and tools such as NVIDIA CUDA-Q, Helios bridges classical and quantum computing more seamlessly than ever, making hybrid quantum-classical development simple, natural, and accessible, and establishing Helios as the most programmable, general-purpose quantum computer ever built.
The Most Logical Path to Fault Tolerance
While everyone else is promising fault-tolerance, we’re delivering it. We are the only company to demonstrate a fully universal fault-tolerant gate set, we’ve demonstrated more codes than anyone else, and our logical fidelities are the best in class.
Now, with 98 physical qubits, we’ve been able to make 94 logical qubits, fully entangled in one of the largest GHZ states ever recorded. We did this with better than break-even fidelity, meaning they outperform physical qubits running the same algorithm. Built on our Iceberg code, published last year in Nature Physics, these logical qubits achieve the industry’s highest encoding efficiency, needing only two ancilla qubits per code block, or roughly a 1:1 physical-to-logical qubit ratio.
With 50 error-detected logical qubits, Helios achieved better than break-even performance, running the largest encoded simulation of quantum magnetism to date—an exceptional example of how users can leverage efficient encodings. This range and flexibility let users tailor the encoding rate to their application: fewer logical qubits deliver higher fidelity for less complex tasks, while larger sets enable more complex simulations.
Helios also produced 48 fully error-corrected logical qubits at a remarkable 2:1 encoding rate, a ratio thought impossible just a few years ago. This super high encoding rate stands in stark contrast to other notable demonstrations from industry peers. For example, the demonstration linked in the previous sentence would need a whopping 4800 qubits to make 48 logical qubits. Our 2:1 encoding rate was achieved through a clever technique called code concatenation, a breakthrough that supports single-shot error correction, transversal logic, and full parallelization—all at 99.99% state preparation and measurement fidelity.
To extend this performance at scale, all future Quantinuum systems—starting with Helios—will integrate real-time decoding using NVIDIA Grace Hopper GPUs, treating decoding as a dynamic computational process rather than a static lookup. Errors can be corrected as computations run without slowing the logical clock rate. Combined with Guppy, NVIDIA CUDA-Q, and NVQLink, this infrastructure forms the foundation for fault-tolerant, real-time quantum computation, delivering immediate quantum advantage in the near term and a clear path to scalable error-corrected computing.
We remain the only company to perform a fully universal fault-tolerant gate set, with more error-correcting codes and higher logical fidelities than any other company.
Helios is ready to drive practical, commercial quantum applications across industries. Its unprecedented fidelity, scalability, and programmability give users the tools to tackle problems that were previously out of reach. This is just the beginning, and we look forward to seeing what users and companies will achieve with it.
Typically, Quantum Error Detection (QED) is viewed as a short-term solution—a non-scalable, stop-gap until full fault tolerance is achieved at scale.
That’s just changed, thanks to a serendipitous discovery made by our team. Now, QED can be used in a much wider context than previously thought. Our team made this discovery while studying the contact process, which describes things like how diseases spread or how water permeates porous materials. In particular, our team was studying the quantum contact process (QCP), a problem they had tackled before, which helps physicists understand things like phase transitions. In the process (pun intended), they came across what senior advanced physicist, Eli Chertkov, described as “a surprising result.”
While examining the problem, the team realized that they could convert detected errors due to noisy hardware into random resets, a key part of the QCP, thus avoiding the exponentially costly overhead of post-selection normally expected in QED.
To understand this better, the team developed a new protocol in which the encoded, or logical, quantum circuit adapts to the noise generated by the quantum computer. They quickly realized that this method could be used to explore other classes of random circuits similar to the ones they were already studying.
The team put it all together on System Model H2 to run a complex simulation, and were surprised to find that they were able to achieve near break-even results, where the logically encoded circuit performed as well as its physical analog, thanks to their clever application of QED. Ultimately, this new protocol will allow QED codes to be used in a scalable way, saving considerable computational resources compared to full quantum error correction (QEC).
Researchers at the crossroads of quantum information, quantum simulation, and many-body physics will take interest in this protocol and use it as a springboard for inventing new use cases for QED.
Stay tuned for more, our team always has new tricks up their sleeves.
Learn mode about System Model H2 with this video: