

AI + quantum computing: Quantinuum, NVIDIA, and Pfizer have combined transformer-based generative AI with quantum computing to automatically generate high-quality quantum chemistry circuits more efficiently than traditional optimization methods.
Practical pharma impact: The approach was used to prepare molecular ground states and validated on Quantinuum’s Helios hardware, demonstrating a path toward larger-scale computational chemistry and drug discovery.
Long-term vision: The team aims to build quantum foundation models that learn from increasingly complex quantum data, eventually enabling AI to design circuits for molecules too large for classical simulation.
Quantum computing has long promised a future that expands what we can do with compute — for example, in molecular simulation, materials discovery, or pharmaceuticals development. But between that promise and practical utility sits a stubborn bottleneck: quantum state preparation.
To run any algorithm on a quantum computer, you must first put the qubits in the right starting state. Think of it like setting up a Rube Goldberg machine- except in this case, you’re not sure exactly which initial setup will give you the results you want. This is what makes quantum state preparation so important: your choice of initial state dictates the accuracy and cost of the rest of the calculation.
We teamed up with NVIDIA and Pfizer to tackle this problem, with an eye towards developing meaningful industrial workflows. The result is a new generative quantum AI framework, called ADAPT-GQE, which we consider to be a canonical instance of GenQAI. ADAPT-GQE uses quantum data to train transformer models that ultimately synthesize quantum chemistry circuits faster, with better outcomes, in a sort of ‘virtuous cycle’.
Ultimately, this means we have developed a new interface between quantum computing and AI. By treating quantum circuit generation as a language modelling problem, we now have a system that can generate high-quality ground-state preparation circuits - with comparable or improved state preparation accuracy.
The goal of computational chemistry is to learn about chemical properties without performing expensive, time-consuming, and sometimes dangerous “wet-lab” experiments.
In principle, you can replace the majority of your physical experiments with computer simulations, saving billions of dollars and years of time.
In reality, computational chemistry is very tricky. To accurately simulate a chemical inside of a computer, you have to build it from the ground up. You start with a collection of atoms (in the case of imipramine, you have 19 Carbon atoms, 24 Hydrogen atoms, and 2 Nitrogen atoms). Then, like Nature’s ‘lego’, you assemble those atoms into a molecule: you set bond lengths, strengths, angles, interactions, and so on.
This is not straightforward: a single molecule can exist in many forms; with different angles, rotations, etc. We will call these different forms ‘conformations’.
Then, to actually estimate chemical properties, or to explore chemical reaction pathways, you have to reproduce the detailed physics that goes on at the atomic level: take your chosen conformation then figure out how each orbital is occupied, how the electrons are interacting with each other or the atomic nuclei, how is the addition of heat or a catalyst going to affect things.... it gets complicated, quickly.
Despite all this, computational chemistry is a powerhouse in pharmaceutical development. Right now, pharmaceutical companies save money and time by simulating as much as they can on computers, avoiding time consuming and expensive laboratory experiments. However, even with ~50 years of development, the existing classical methods have very real limitations.
This is where quantum computing comes in: this new computational paradigm can elide those limitations because it has many of the “hard parts” (like superposition or entanglement) natively encoded. Used correctly, quantum computing promises to break old barriers, further improving margins for pharma companies across the globe while contributing to meaningful, impactful, discoveries.
While quantum computational chemistry is one of the strongest candidates for near-term quantum advantage, current hardware is still in the earlier stages of development. With limited qubits and error rates, algorithm designers need to make every gate count, keep circuits shallow, and be able to tolerate some level of noise.
This is where generative AI enters the picture.
Instead of hand-designing chemistry circuits and laboriously experimenting to see how well they run, there is another idea: what if we trained an AI to solve the problems that quantum computational chemistry faces?
Using this approach, not only can we save time and resources; but we can shorten the timeline to realize practical results. With better state prep and other circuits, applications that were once considered far in the future come into view.
Our first attempt at this is called ADAPT-GQE. The central idea behind ADAPT-GQE is deceptively simple: instead of laboriously searching for good quantum circuits from scratch, train a transformer model to generate them directly.
Importantly, the framework is model-agnostic, which we showed by deploying it on complementary transformer architectures - Nemotron (a pretrained LLM) and Gemma (trained from scratch).
The initial goal here is to find the ‘ground state’ of the molecule imipramine (this is the electronic state with the smallest amount of energy stored inside it). To do this, you have to find the right ‘state preparation circuit’, as described above.
Until now, a leading method for finding the ground state with quantum computers was the ‘Variational Quantum Eigensolver (VQE)’, a hybrid quantum-classical approach. The VQE process starts with a ‘guess’ circuit for a particular conformation of the molecule. The quantum computer runs the circuit to measure the associated energy of the molecule. This result is fed back into a classical optimizer that then tweaks the circuit parameters, hopefully resulting in one with a lower molecular energy. This loop repeats until a minimum energy is found.
Unfortunately, VQE has a few severe limitations that make it infeasible for widespread use. The recently proposed ADAPT-VQE was a crucial step forward meant to address some of the issues with “plain” VQE. In ADAPT-VQE, instead of starting with a guess for the initial circuit, the process builds a circuit in steps by selecting operators from a pool(typically using gradient information) and optimizing. This approach can be more effective, but unfortunately still grows too large too quickly.
This is where the joint team jumped in.
Combining the best of all worlds, the team’s new framework, ADAPT-GQE, combines AI with the ADAPT-VQE to create something entirely new – and something that, so far, is a scalable, hardware-validated pathway toward automated quantum circuit synthesis.
First, transformers (in this case, Nemotron and Gemma) are trained via supervised fine-tuning on ADAPT-VQE data. In this way, the old method isn’t thrown away but is instead treated as a high-quality data-producing “oracle”.
Then, once the transformer has been initially trained, it defines a distribution over circuits, each one with some probability of corresponding to the ground state. This distribution can be used in a fine-tuning loop, for example, reinforcement learning. In reinforcement learning, the framework takes a circuit from that distribution, runs it, and measures the energy. It feeds the results back into the transformer, which adjusts its distribution. Over time, the model learns to prioritize circuits that prepare increasingly accurate ground states.
Crucially, reinforcement learning allows the system to surpass its original training data instead of merely imitating it. The model is no longer acting as a compressed lookup table for ADAPT-VQE. It begins exploring novel circuit configurations that may outperform the teacher algorithm itself. This is one of the most important conceptual shifts in the project.
In this case, instead of running all the initial circuits on Quantinuum’s Helios, the reinforcement learning circuits were run using NVIDIA accelerated computing and the CUDA-Q platform, simulating a quantum processor.
Finally, once the transformers are optimized via reinforcement learning, the best resulting circuits are validated for accuracy and feasibility, by running them using InQuanto and Nexus on Quantinuum’s newest hardware, Helios. With InQuanto v5.2, users can now interface directly with both the Helios quantum computer and the Selene quantum emulator through Nexus.
This powerful combination of InQuanto and Nexus enabled the execution one of the largest AI-generated quantum chemistry circuits to date on a quantum computer; helping to turn the promise of quantum computing into a practical tool for pharmaceutical development.
Looking farther in the future, the researchers envision something much larger than a single molecular benchmark.
For bigger and more complex molecules, ADAPT-VQE won’t work in the first place as the initial training “oracle”. In addition, the molecular energy calculations used in the reinforcement learning grow too large for classical systems simulating quantum computers, so the quantum processor becomes essential.
Luckily, this is not a problem. The ultimate goal of the ADAPT-GQE framework is to develop a “curriculum” for the transformers. This means instead of re-training them for every new molecule, you instead keep what you already learned, and expand your knowledge from there.
By initially teaching it on molecules that are smaller, and that can be fully simulated, you ensure it learns on good data that can be double checked using known methods. From there, you can carefully build up the complexity to see how the transformer learns. Eventually, you hope to train it on molecules that can’t be simulated classically, using purely quantum data, all the while getting closer to the complexity levels you’re chasing.
This penultimate result is called a ‘foundation model’, which is a massive AI neural network trained on vast, broad datasets that can be adapted to a wide variety of downstream tasks. In this case, the team is building the very ‘foundations’ of a model that can solve the ‘electronic structure problem’, which is the core computational challenge lying at the heart of quantum (and classical) computational chemistry.
What makes this work particularly interesting is that it treats quantum circuit generation as a language modeling problem: circuits become sequences, transformers learn distributions over those sequences, and reinforcement learning optimizes them against physical reward functions.
The result is an AI system capable of proposing quantum circuits that were never explicitly programmed by humans.
That does not mean generative AI is replacing physics or chemistry. Instead, it is becoming a new interface layer for navigating unimaginably large search spaces that traditional optimization methods struggle to explore efficiently.
For quantum chemistry, that could become transformative.
If successful, frameworks like ADAPT-GQE may eventually allow researchers to synthesize useful quantum circuits for molecular systems too large for classical computation, accelerating everything from materials discovery to pharmaceutical design.
The broader implication is difficult to ignore: foundation models may eventually extend beyond language, images, and code — and into the fabric of physical reality itself.
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 next era of computing will be shaped by bringing different computing resources together, and quantum has earned a place in that conversation. At SC26 in Chicago, Quantinuum is showing how we’re bringing that hybrid integration to life today.
Consider a researcher studying a molecule. Classical computers prepare the problem, a quantum computer contributes a calculation, and classical computers analyze the results. Those steps may run on different machines, in different locations. Connecting them into a coordinated workflow is essential to making the application useful.
Quantinuum Nexus is our developer platform that helps manage those connections. It brings quantum programming, testing and execution into one environment and provides the foundation to coordinate quantum calculations with the surrounding classical work. Researchers can focus on the scientific problem without writing custom software to manage every handoff between systems.
Making these connections easier helps more people build quantum applications, including those whose expertise lies in the scientific and industrial problems they want to solve.
A researcher studying a molecule or material needs more than access to a quantum processor. A hybrid application can involve classical preparation, HPC simulation, quantum execution and post-processing. Each stage needs particular resources and may depend on results from an earlier calculation.
As the application grows, coordinating those steps becomes a development challenge. Developers need to specify which tasks can run independently, which must wait for results, and where each task should execute. The resources also operate on different timescales. A quantum job may be waiting in a queue or running remotely. During that time, valuable HPC resources should remain available for other work.
Nexus’ Hybrid Workflow Runtime helps address this challenge. Developers express how the quantum and classical tasks depend on one another, and the runtime coordinates their execution across the relevant resources. Its asynchronous approach allows work to progress without requiring every resource to remain active or blocked while another task completes. This connects quantum job submission to the surrounding classical application, reducing the need for developers to write bespoke logic.
Our work with RIKEN illustrates the scientific workflows this coordination can support.
Calculating how electrons behave in a molecule is central to understanding its properties. These electronic-structure calculations help researchers investigate questions relevant to chemistry and materials science, but they can become increasingly demanding as the problem grows.
RIKEN’s quantum residual exploration method combines quantum and classical contributions to that calculation. Quantum measurements suggest additional electronic configurations—possible arrangements of electrons—which classical routines assess and select for the final calculation.
In a demonstration involving a stretched 12-atom hydrogen chain, the quantum-generated proposals lowered the calculated energy compared with classical-only selection using the same final number of configurations. This improved the energy estimate within that specific comparison.
Our developer tools and platform connected the stages of the application. Guppy expressed the quantum logic, Hybrid Workflow Runtime (Tierkreis) coordinated the quantum and classical tasks, and Nexus provided the submission interface for quantum execution. Researchers could develop the application without manually coordinating every handoff.
As hybrid applications develop, their workflows need to adapt to different resources, tools and operating practices.
That makes interoperability a design requirement for Nexus. Developers should be able to choose resources suited to their application. Computing centers need flexibility to build on existing investments and evolve their infrastructure over time.
HPC scheduling is one example. Quantinuum supports connections to workload managers including Slurm, PBS, and PJSub, enabling developers to build around the scheduling infrastructure at their computing center.
With Slurm, Hybrid Workflow Runtime can submit classical tasks to a Slurm-managed environment while quantum jobs are submitted through Nexus. The runtime coordinates the task dependencies, while Slurm allocates the compute resources those tasks need, across both HPC and now QPUs as well given the recent expansion to Slurm announced by NVIDIA to include quantum resources.
Interoperability extends beyond scheduling. A computing center may combine workflow tools, resource-management software and quantum hardware from several providers. Defined interfaces help those components work together as the environment changes.
One example is the Open Quantum-HPC Software Ecosystem, or OpenQSE, a community effort led by Oak Ridge National Laboratory. Its reference architecture defines boundaries across the hybrid computing stack to help different implementations work together. Within that architecture, Nexus' Hybrid Workflow Runtime sits at the application workflow layer, above scheduling and resource-access software. Interfaces such as QRMI and QDMI provide connections toward quantum systems.
Our SC26 demonstration in Chicago illustrates this division of responsibilities. Hybrid Workflow Runtime orchestrates the application, assigning classical chemistry tasks to a shared Slurm executor with an explicit resource profile. The runtime coordinates task dependencies, while Slurm handles placement and hybrid compute resource allocation. The workflow can then interact with QFw to reach quantum execution targets.
Explore the SC26 demonstration here.
The applications researchers develop today also help build the expertise needed for increasingly capable quantum computers.
The Nexus fault-tolerance roadmap is intended to support that transition. Supported error-correction encoding, emulation and execution on Helios already provide a starting point. Future releases aim to simplify more of the work involved in developing and deploying quantum applications as the hardware advances.
Through Nexus, we are connecting quantum programming with hybrid deployment. Our goal is for the applications, workflows and expertise users develop today to provide a foundation they can carry forward into the fault-tolerant era.
That preparation involves learning how to formulate a useful problem, test an approach, evaluate results and coordinate the computing resources needed to run the application. Making those steps easier helps more people contribute, including researchers whose expertise lies in the scientific and industrial problems they want to solve.
At SC26, we’re bringing these pieces together: quantum hardware, developer tools, and the connections that make hybrid applications possible within existing HPC environments.
You can visit Quantinuum at Booth 4838 in the Quantum Junction, November 17–19 at McCormick Place, Chicago. See our hybrid workflow demonstration and join more than 15 booth theater sessions with Quantinuum experts and partners including NVIDIA and Oracle.
Bring your questions about connecting quantum to your HPC environment, developing hybrid applications, and preparing for what comes next.
Book time with our experts, explore the Nexus roadmap, or get started with Nexus.
Quantinuum is heading to SC26 in Chicago, and we'd love to see you at Booth 4838 in the Quantum Junction. Stop by during exhibit hours to catch partner theater sessions throughout the day, see what's new with Helios and Nexus, and talk with our team about how quantum computing is expanding its footprint in the HPC datacenter, including hybrid quantum-classical applications.
Join our experts and partners at our kiosks and for short sessions in our theater running throughout the day. Topics include:
Need a break from the show floor? Our barista will be serving up coffee, and we'll have fresh cookies every afternoon. Grab a cup, pull up a seat for a theater session, and chat with the team.
Our experts are also speaking in the SC26 technical program. Add these sessions to your schedule (all times CST):
9:45 – 10:00am | Room: W178a
Quantum-HPC Workflows Across Multiple Quantum Computing Platforms: Two Case Studies – part of the 2nd International Workshop for Software Frameworks and Workload Management on Quantum-HPC Ecosystems
Presenting: Philipp Seitz
3:30 – 4:00pm | Room: W179a
Software for Large Scale Application – part of the 2nd Annual Workshop on Large-Scale Quantum-Classical Computing (LSQCC)
Presenting: Juan Pederson
3:30 – 4:00pm | Room: W181abc
Bringing Quantum Computing to the HPC Datacenter
Presenting: Kentaro Yamamoto
Note: Schedule subject to changes.
Whether you're exploring quantum for the first time or planning how to integrate it with your HPC environment, our team is ready to talk. Find us at Booth 4838 in the Quantum Junction, November 16–19.
Quantinuum has been recognized as a “Market Shaper” in Gartner’s first Emerging Market Quadrant for Quantum Computing. According to Gartner, Market Shapers combine a high potential to transform the market with strong execution capabilities.
You can access the report here.
While this recognition is welcome and validates much of what we have contributed to the quantum computing space, we believe there is compelling evidence that supports an even stronger leadership position.
The distinction matters for customers deciding where to strategically invest their time, expertise and resources. Quantum leadership requires reliable computation at useful speed, software that enables application development across hardware generations, and the capacity to deliver at scale. Across these measures, our proven results, growing adoption and long-term customer commitments demonstrate Quantinuum’s leadership.
Quantinuum’s position has been repeatedly recognized by others in the industry as well, including experts like André M. König, CEO of Global Quantum Intelligence (GQI).
According to André: “Traditional IT analyst frameworks often struggle with quantum because they weigh near-term service wrappers on the same scale as deep-tech physics. Most of what this industry publishes is a slideware roadmap. Quantinuum’s recent error-correction milestones - reaching 99.9995% logical memory fidelity on Helios - were demonstrated on a commercially deployed system with the data open to inspection. For enterprise buyers and sovereign programs trying to separate engineering reality from marketing ambition, demonstrated logical performance is the only metric that matters today.”
Quantinuum’s progress places us at the forefront of scalable, fault-tolerant computing. We bring together the three capabilities essential to achieving it: accurate hardware, an architecture that scales efficiently, and error correction that improves reliability while managing resource requirements.
In the third quarter, we introduced an end-to-end logical architecture on Helios that combines tailored AI decoders, all-to-all qubit connectivity, high-fidelity physical operations and hardware-level error correction. Two different error-correcting codes running within that architecture on the same hardware delivered back-to-back world records in September and October, reinforcing what we believe is the industry’s most credible path to universal, fully fault-tolerant quantum computing.
October’s result reached 99.9995% logical memory fidelity, equivalent to five errors in one million operations. That error rate was ten times lower than our September record and 100 times lower than the previous best published result across other hardware modalities, achieved by Google.
The results also validated the scalability of our architecture. Each two-step increase in error-correction code distance reduced logical memory errors by approximately twelvefold, compared with roughly twofold in Google’s published superconducting result. We estimate this stronger suppression points to roughly 30 times fewer physical qubits needed to reach target logical error rates in that comparison—a significant advantage for scaling.
These results build on years of advancing our architecture and delivering successive systems, from H1 through Helios. That record gives customers a concrete basis for confidence in our roadmap toward Sol and Apollo.
Leadership also means helping customers understand which performance measures matter. As our Chief Scientist, Harry Buhrman, explains in Fidelity First in the Quantum Race, faster gates do not necessarily produce faster useful answers. Higher fidelity can reduce the additional qubits and operations needed for error correction, shortening the time to a reliable result.
QUOPS, developed by Sandia National Laboratories with input from Quantinuum and NVIDIA, helps customers evaluate both computational capability and speed. In the first published study, Helios achieved the highest computational capability score of the four systems assessed—a measure of circuit size—at 1,504, or 1,824 with leakage post-selection. Other systems achieved higher computational rates, but on smaller circuits.
These dimensions must be considered together to understand what a system can reliably deliver. We are working toward an industry standard around QUOPS so customers can compare systems and make better-informed purchasing decisions.
Our leadership extends to the software customers need to put increasingly capable hardware to work.
Quantinuum Nexus brings quantum application development and deployment into one environment, enabling developers and error-correction experts to advance their work together. Its multigenerational roadmap aligns software capabilities with successive hardware generations through Apollo and beyond, helping customers carry forward the applications, workflows and expertise they build today.
We are already delivering against that roadmap. The September release extended our Guppy programming language with tools for programming logical qubits and applying supported error-correction codes, reducing the need to rewrite application logic for each code. The planned December release will expand hybrid workflows, allowing applications to respond to quantum results as they become available, alongside additional codes and execution analytics.
Adoption provides another measure of progress. In the third quarter, 215 organizations were building on Nexus, up nearly 20% from the previous quarter. Total job submissions increased tenfold in the twelve months through August 2026 compared with the preceding twelve-month period.
Our approach also connects quantum computing with high-performance computing and AI. At RIKEN in Japan, Quantinuum’s Reimei quantum computer works alongside the Fugaku supercomputer and ROQUO GPU supercomputer. Researchers are exploring materials and biomolecular chemistry through workflows combining quantum calculations, classical simulation and AI.
Customer adoption and commitments provide another concrete measure of our leadership.
Our collaborations span financial services, drug development, materials science and logistics, focusing research on specific customer problems. With Johnson & Johnson, we are applying quantum computing and AI to the creation of new cancer-therapy materials. With GE Vernova, we are advancing research into superconducting materials for smaller, lighter and more efficient electrical infrastructure. Our collaboration with Abu Dhabi Maritime Academy and Quincus explores quantum, AI and high-performance computing for maritime and logistics applications.
Customers are also committing across hardware generations. In the third quarter, Quantinuum secured its largest cloud agreement: a multiyear extension with a leading global financial institution that includes committed access to our systems through Apollo using Nexus, supporting fault-tolerant application development and deployment across multiple core business functions.
This is in addition to our Oracle partnership. Plans to deploy Helios within an Oracle Cloud Infrastructure AI data center will place quantum alongside GPU and high-performance computing resources. Through Oracle’s Quantum AI Compute beta program, selected organizations can already explore hybrid quantum and AI applications.
In the third quarter, we also signed an agreement with another leading hyperscaler serving millions of developers and enterprises worldwide, further expanding access to our hardware and software.
Quantinuum is also building and scaling manufacturing partnerships.
Our longstanding partnerships with GlobalFoundries, Infineon, Honeywell Aerospace and Monarch Quantum bring manufacturing expertise into critical-component design early, helping reduce complexity, improve reliability and make production more repeatable. Our finalized CHIPS R&D award further supports these efforts; Quantinuum was the only trapped-ion company selected for the award.
We believe Quantinuum’s demonstrated performance, software capabilities and customer commitments establish the quantum computing industry’s strongest leadership position.
Our focus is on extending that lead through execution: advancing toward Sol and Apollo, expanding adoption of Nexus, and delivering the capabilities customers need to put quantum computing to work.
This post contains forward-looking statements, including statements about Quantinuum's technology and product roadmap (including Sol and Apollo), planned software releases and platform expansion, the scalability and performance of our architecture, and anticipated customer adoption. Words such as "will," "plan," "expect," "believe," "roadmap" and similar expressions identify forward-looking statements. These statements reflect our current expectations and assumptions and are subject to risks and uncertainties, including technical, development, manufacturing, supply chain, partner and market risks, that could cause actual results to differ materially. Forward-looking statements speak only as of the date of this post, and Quantinuum undertakes no obligation to update them, except as required by law.