Quantum Computers Will Make AI Better
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.
- The Sanger Institute illustrates the value of quantum computing to genomics research
- Quantinuum supports developments in a field that promises to deliver a profound and positive societal impact
Twenty-five years ago, scientists accomplished a task likened to a biological moonshot: the sequencing of the entire human genome.
The Human Genome Project revealed a complete human blueprint comprising around 3 billion base pairs, the chemical building blocks of DNA. It led to breakthrough medical treatments, scientific discoveries, and a new understanding of the biological functions of our body.
Thanks to technological advances in the quarter-century since, what took 13 years and cost $2.7 billion then can now be done in under 12 minutes for a few hundred dollars. Improved instruments such as next-generation sequencers and a better understanding of the human genome – including the availability of a “reference genome” – have aided progress, alongside enormous advances in algorithms and computing power.
But even today, some genomic challenges remain so complex that they stretch beyond the capabilities of the most powerful classical computers operating in isolation. This has sparked a bold search for new computational paradigms, and in particular, quantum computing.
Quantum Challenge: Accepted
The Wellcome Leap Quantum for Bio (Q4Bio) challenge is pioneering this new frontier. The program funds research to develop quantum algorithms that can overcome current computational bottlenecks. It aims to test the classical boundaries of computational genetics in the next 3-5 years.
One consortium – led by the University of Oxford and supported by prestigious partners including the Wellcome Sanger Institute, the Universities of Cambridge, Melbourne, and Kyiv Academic University – is taking a leading role.
“The overall goal of the team’s project is to perform a range of genomic processing tasks for the most complex and variable genomes and sequences – a task that can go beyond the capabilities of current classical computers” – Wellcome Sanger Institute press release, July 2025
Selecting Quantinuum
Earlier this year, the Sanger Institute selected Quantinuum as a technology partner in their bid to succeed in the Q4Bio challenge.
Our flagship quantum computer, System H2, has for many years led the field of commercially available systems for qubit fidelity and consistently holds the global record for Quantum Volume, currently benchmarked at 8,388,608 (223).
In this collaboration, the scientific research team can take advantage of Quantinuum’s full stack approach to technology development, including hardware, software, and deep expertise in quantum algorithm development.
“We were honored to be selected by the Sanger Institute to partner in tackling some of the most complex challenges in genomics. By bringing the world’s highest performing quantum computers to this collaboration, we will help the team push the limits of genomics research with quantum algorithms and open new possibilities for health and medical science.” – Rajeeb Hazra, President and CEO of Quantinuum
Quantum for Biology
At the heart of this endeavor, the consortium has announced a bold central mission for the coming year: to encode and process an entire genome using a quantum computer. This achievement would be a potential world-first and provide evidence for quantum computing’s readiness for tackling real-world use cases.
Their chosen genome, the bacteriophage PhiX174, carries symbolic weight, as its sequencing earned Fred Sanger his second Nobel Prize for Chemistry in 1980. Successfully encoding this genome quantum mechanically would represent a significant milestone for both genomics and quantum computing.
Sooner than many expect, quantum computing may play an essential role in tackling genomic challenges at the very frontier of human health. The Sanger Institute and Quantinuum’s partnership reminds us that we may soon reach an important step forward in human health research – one that could change medicine and computational biology as dramatically as the original Human Genome Project did a quarter-century ago.
“Quantum computational biology has long inspired us at Quantinuum, as it has the potential to transform global health and empower people everywhere to lead longer, healthier, and more dignified lives.” – Ilyas Khan, Founder and Chief Product Officer of Quantinuum
Every year, The IEEE International Conference on Quantum Computing and Engineering – or IEEE Quantum Week – brings together engineers, scientists, researchers, students, and others to learn about advancements in quantum computing.
This year’s conference from August 31st – September 5th, is being held in Albuquerque, New Mexico, a burgeoning epicenter for quantum technology innovation and the home to our new location that will support ongoing collaborative efforts to advance the photonics technologies critical to furthering our product development.
Throughout IEEE Quantum Week, our quantum experts will be on-site to share insights on upgrades to our hardware, enhancements to our software stack, our path to error correction, and more.
Meet our team at Booth #507 and join the below sessions to discover how Quantinuum is forging the path to fault-tolerant quantum computing with our integrated full-stack.
September 2nd
Quantum Software Workshop
Quantum Software 2.1: Open Problems, New Ideas, and Paths to Scale
1:15 – 2:10pm MDT | Mesilla
We recently shared the details of our new software stack for our next-generation systems, including Helios (launching in 2025). Quantinuum’s Agustín Borgna will deliver a lighting talk to introduce Guppy, our new, open-source programming language based on Python, one of the most popular general-use programming languages for classical computing.
September 3rd
PAN08: Progress and Platforms in the Era of Reliable Quantum Computing
1:00 – 2:30pm MDT | Apache
We are entering the era of reliable quantum computing. Across the industry, quantum hardware and software innovators are enabling this transformation by creating reliable logical qubits and building integrated technology stacks that span the application layer, middleware and hardware. Attendees will hear about current and near-term developments from Microsoft, Quantinuum and Atom Computing. They will also gain insights into challenges and potential solutions from across the ecosystem, learn about Microsoft’s qubit-virtualization system, and get a peek into future developments from Quantinuum and Microsoft.
BOF03: Exploring Distributed Quantum Simulators on Exa-scale HPC Systems
3:00 – 4:30pm MDT | Apache
The core agenda of the session is dedicated to addressing key technical and collaborative challenges in this rapidly evolving field. Discussions will concentrate on innovative algorithm design tailored for HPC environments, the development of sophisticated hybrid frameworks that seamlessly combine classical and quantum computational resources, and the crucial task of establishing robust performance benchmarks on large-scale CPU/GPU HPC infrastructures.
September 4th
PAN11: Real-time Quantum Error Correction: Achievements and Challenges
1:00 – 2:30pm MDT | La Cienega
This panel will explore the current state of real-time quantum error correction, identifying key challenges and opportunities as we move toward large-scale, fault-tolerant systems. Real-time decoding is a multi-layered challenge involving algorithms, software, compilation, and computational hardware that must work in tandem to meet the speed, accuracy, and scalability demands of FTQC. We will examine how these challenges manifest for multi-logical qubit operations, and discuss steps needed to extend the decoding infrastructure from intermediate-scale systems to full-scale quantum processors.
September 5th
Keynote by NVIDIA
8:00 – 9:30am MDT | Kiva Auditorium
During his keynote talk, NVIDIA’s Head of Quantum Computing Product, Sam Stanwyck, will detail our partnership to fast-track commercially scalable quantum supercomputers. Discover how Quantinuum and NVIDIA are pushing the boundaries to deliver on the power of hybrid quantum and classical compute – from integrating NVIDIA’s CUDA-Q Platform with access to Quantinuum’s industry-leading hardware to the recently announced NVIDIA Quantum Research Center (NVAQC).
Featured Research at the IEEE Poster Session:
Visible Photonic Component Development for Trapped-Ion Quantum Computing
September 2nd from 6:30 - 8:00pm MDT | September 3rd from 9:30 - 10:00am MDT | September 4th from 11:30 - 12:30pm MDT
Authors: Elliot Lehman, Molly Krogstad, Molly P. Andersen, Sara Cambell, Kirk Cook, Bryan DeBono, Christopher Ertsgaard, Azure Hansen, Duc Nguyen, Adam Ollanik, Daniel Ouellette, Michael Plascak, Justin T. Schultz, Johanna Zultak, Nicholas Boynton, Christopher DeRose,Michael Gehl, and Nicholas Karl
Scaling Up Trapped-Ion Quantum Processors with Integrated Photonics
September 2nd from 6:30 - 8:00pm MDT and 2:30 - 3:00pm MDT | September 4th from 9:30 - 10:00am MDT
Authors: Molly Andersen, Bryan DeBono, Sara Campbell, Kirk Cook, David Gaudiosi, Christopher Ertsgaard, Azure Hansen, Todd Klein, Molly Krogstad, Elliot Lehman, Gregory MacCabe, Duc Nguyen, Nhung Nguyen, Adam Ollanik, Daniel Ouellette, Brendan Paver, Michael Plascak, Justin Schultz and Johanna Zultak
Research Collaborations with the Local Ecosystem
In a partnership that is part of a long-standing relationship with Los Alamos National Laboratory, we have been working on new methods to make quantum computing operations more efficient, and ultimately, scalable.
Learn more in our Research Paper: Classical shadows with symmetries
Our teams collaborated with Sandia National Laboratories demonstrating our leadership in benchmarking. In this paper, we implemented a technique devised by researchers at Sandia to measure errors in mid-circuit measurement and reset. Understanding these errors helps us to reduce them while helping our customers understand what to expect while using our hardware.
Learn more in our Research Paper: Measuring error rates of mid-circuit measurements
From machine learning to quantum physics, tensor networks have been quietly powering the breakthroughs that will reshape our society. Originally developed by the legendary Nobel laureate Roger Penrose, they were first used to tackle esoteric problems in physics that were previously unsolvable.
Today, tensor networks have become indispensable in a huge number of fields, including both classical and quantum computing, where they are used everywhere from quantum error correction (QEC) decoding to quantum machine learning.
In this latest paper, we teamed up with luminaries from the University of British Columbia, California Institute of Technology, University of Jyväskylä, KBR Inc, NASA, Google Quantum AI, NVIDIA, JPMorgan Chase, the University of Sherbrooke, and Terra Quantum AG to provide a comprehensive overview of the use of tensor networks in quantum computing.
Standing on the shoulders of giants
Part of what drives our leadership in quantum computing is our commitment to building the best scientific team in the world. This is precisely why we hired Dr. Reza Haghshenas, one of the world’s leading experts in tensor networks, and a co-author on the paper.
Dr. Haghshenas has been researching tensor networks for over a decade across both academia and industry. Dr. Haghshenas did postdoctoral work under Professor Garnet Chan at Caltech, a leading figure in the use of tensor networks for quantum physics and chemistry.
“Working with Dr. Garnet Chan at Caltech was a formative experience for me”, remarked Dr. Haghshenas. “While there, I contributed to the development of quantum simulation algorithms and advanced classical methods like tensor networks to help interpret and simulate many-body physics.”
Since joining Quantinuum, Dr. Haghshenas has led projects that bring tensor network methods into direct collaboration with experimental hardware teams — exploring quantum magnetism on real quantum devices and helping demonstrate early signs of quantum advantage. He also contributes to widely used simulation tools like QUIMB, helping the broader research community access these methods.
Dr. Haghshenas’ work sits in a broad and vibrant ecosystem exploring novel uses of tensor networks. Collaborations with researchers like Dr. Chan at Caltech, and NVIDIA have brought GPU-accelerated tools to bear on the forefront of applying tensor networks to quantum chemistry, quantum physics, and quantum computing.
A powerful simulation tool
Of particular interest to those of us in quantum computing, the best methods (that we know of) for simulating quantum computers with classical computers rely on tensor networks. Tensor networks provide a nice way of representing the entanglement in a quantum algorithm and how it spreads, which is crucial but generally quite difficult for classical algorithms. In fact, it’s partly tensor networks’ ability to represent entanglement that makes them so powerful for quantum simulation. Importantly, it is our in-house expertise with tensor networks that makes us confident we are indeed moving past classical capabilities.
A theory of evolution
Tensor networks are not only crucial to cutting-edge simulation techniques. At Quantinuum, we're working on understanding and implementing quantum versions of classical tensor network algorithms, from quantum matrix product states to holographic simulation methods. In doing this, we are leveraging decades of classical algorithm development to advance quantum computing.
A topic of growing interest is the role of tensor networks in QEC, particularly in a process known as decoding. QEC works by encoding information into an entangled state of multiple qubits and using syndrome measurements to detect errors. These measurements must then be decoded to identify the specific error and determine the appropriate correction. This decoding step is challenging—it must be both fast (within the qubit’s coherence time) and accurate (correctly identifying and fixing errors). Tensor networks are emerging as one of the most effective tools for tackling this task.
Looking forward (and backwards, and sideways...)
Tensor networks are more than just a powerful computational tool — they are a bridge between classical and quantum thinking. As this new paper shows, the community’s understanding of tensor networks has matured into a robust foundation for advancing quantum computing, touching everything from simulation and machine learning to error correction and circuit design.
At Quantinuum, we see this as an evolutionary step, not just in theory, but in practice. By collaborating with top minds across academia and industry, we're charting a path forward that builds on decades of classical progress while embracing the full potential of quantum mechanics. This transition is not only conceptual but algorithmic, advancing how we formulate and implement methods utilizing efficiently both classical and quantum computing. Tensor networks aren’t just helping us keep pace with classical computing; they’re helping us to transcend it.