A new report by Hyperion Research and Alice & Bob warns that early fault-tolerant quantum computing (eFTQC) could arrive within five years – reshaping scientific discovery.
In September 2025 Alice & Bob – a global leader in the race for fault-tolerant quantum computing – joined forces with specialist HPC market intelligence firm Hyperion Research to release a new study that could reshape the trajectory of scientific computing.
The report, Seizing Quantum’s Edge: Why and How HPC Should Prepare for eFTQC, presents a clear conclusion: the HPC community must act now to prepare for the arrival of early fault-tolerant quantum computing (eFTQC).
The report says that within the next five to seven years, systems with 100 to 1,000 logical qubits and ultra-low error rates could transform computational science – accelerating workflows beyond the limits of classical supercomputers.
“This is not a distant horizon,” said Bob Sorensen, Senior Vice President and Chief Analyst for Quantum Computing, Hyperion Research.
“Quantum technologies are a pivotal opportunity for the HPC community, offering the potential to significantly accelerate a wide range of critical science and engineering applications in the near term. However, these machines won’t be plug-and-play so HPC centers should begin preparing for integration now.”
The report’s message is urgent yet pragmatic: HPC professionals must invest today in the design of hybrid workflows, software stacks and training programmes that will allow them to exploit eFTQC the moment it becomes available. Those who move early will shape the future of discovery.
The end of classical scaling and the opening for quantum
For decades, HPC performance gains were driven by transistor scaling, energy density improvements and architectural innovation. Yet in the past ten years, the physical limits of transistor size and chip energy capacity have slowed progress. Each new generation of supercomputer is harder, more expensive and less efficient to build.
But the report says at the same time, the quantum horizon has been coming closer. Algorithms once thought to require astronomical quantum resources have become feasible thanks to advances in error correction, algorithm design and qubit encodings. For example, the estimated number of qubits required to run Shor’s algorithm for breaking RSA-2048 encryption has fallen by three orders of magnitude in the past decade.
This dual shift – the slowing of classical HPC scaling and the acceleration of quantum feasibility — has created a critical inflection point. For many workloads, supercomputers are already approaching practical limits while quantum processors are nearing utility. eFTQC is the bridge between the present and the future: a stage where quantum computers are not yet universal machines but are already powerful enough to accelerate meaningful subsets of HPC workloads, the report says.
What eFTQC means
The term “early fault-tolerant quantum computing” refers to machines with hundreds to a thousand logical qubits, operating with logical error rates between 10⁻⁶ and 10⁻¹⁰. Unlike today’s noisy intermediate-scale quantum (NISQ) devices, which can only run approximate algorithms of uncertain value, eFTQC machines will deliver deterministic results through error correction.
The report projects these devices are projected to arrive between 2027 and 2032, depending on hardware progress and vendor roadmaps. While limited in scale, they will be powerful enough to run rigorous algorithms such as quantum phase estimation (QPE) and quantum time evolution — methods that unlock new frontiers in chemistry, materials science and physics, the report says.
“HPC users will see benefits in accuracy, time-to-solution and computational cost as hybrid workflows shift subproblems to quantum processors,” explained Théau Peronnin, CEO, Alice & Bob. “HPC centers that want to lead have to co-design these hybrid workflows with users and vendors, shape efficient infrastructure and deploy prototypes early.”
The promise for scientific domains
Materials Science
Materials science is identified in the report as the first major domain expected to benefit from eFTQC. With just a few hundred logical qubits, quantum computers will be able to solve the Hubbard model in regimes that are classically intractable. This opens the door to breakthroughs in superconductivity, magnetic materials and spin models — fields with enormous industrial implications from lossless power transmission to ultra-efficient computing, the report says.
At leading facilities like the National Energy Research Scientific Computing Center (NERSC) and Los Alamos National Laboratory (LANL), materials simulations already account for up to 20% of workloads. The report says that with eFTQC, those simulations could be accelerated dramatically, freeing resources and unlocking new research questions.
Quantum Chemistry
The next frontier identified in the report is quantum chemistry. Complex molecules with strong electronic correlations, such as FeMoco — the active site of nitrogenase enzymes critical to biological nitrogen fixation — are beyond the reach of classical full configuration interaction methods. Yet eFTQC could make these simulations practical with only a thousand logical qubits. Such capabilities would revolutionise drug discovery, catalysis and energy research, the report says.
“Quantum processors will not replace classical codes wholesale,” said Peronnin. “But they will tackle bottleneck subproblems like ground-state energy calculations that determine the accuracy of larger workflows.”
3. Nuclear Fusion and Energy
Another promising area identified in the report is nuclear fusion research, particularly inertial confinement simulations.
These are computationally intensive problems involving strongly correlated quantum systems and they stand to benefit significantly once eFTQC scales to the upper end of its projected capacity.
By the early 2030s, quantum acceleration could reduce the cost and time required for fusion modeling, advancing humanity’s pursuit of clean, limitless energy, the report says.
4. Fields beyond reach — for now
The report acknowledges that not every domain is suited for eFTQC in the near term. Workloads heavily reliant on solving partial differential equations, such as computational fluid dynamics or structural modeling, require quantum resources far beyond what early devices will provide.
Other fields lack sufficient theoretical groundwork to map their problems onto quantum algorithms.
Still, history suggests new applications will emerge once researchers gain access to the hardware itself, the report says.
Quantifying the Potential Impact
The report provides a systematic analysis of workload distribution at major HPC centers. At NERSC, LANL and DOE leadership facilities, materials science, quantum chemistry and fusion-related applications together account for 30–50% of usage. This suggests that nearly half of the workloads at these institutions could benefit from eFTQC acceleration in the near term.
“The HPC community has always been quick to adopt disruptive architectures — from vector processors to GPUs — and quantum computing is no exception,” said Juliette Peyronnet, US General Manager, Alice & Bob and co-author of the report. “This work is a call to action: centers that begin preparing today will be ready to harness the next major accelerator.”
The quantum frontier is moving fast
One of the most striking findings of the study is the pace at which quantum resource requirements are shrinking. For instance, the number of logical qubits needed to simulate FeMoco has dropped by orders of magnitude in just five years. Similarly, the physical qubit cost of running Shor’s algorithm has plummeted from billions to under a million, thanks to algorithmic improvements and better qubit encodings.
These trends highlight an accelerating timeline. By 2030, some workloads once thought decades away may become feasible on eFTQC machines. This dynamic mirrors the early days of classical computing, when first-use cases often emerged unexpectedly once hardware was deployed, the report says.
Strategic considerations for HPC centers
The report emphasizes that eFTQC will not be a plug-and-play accelerator. Integrating it into HPC environments requires strategic planning across infrastructure, software and workforce development.
The report identifies three pillars as standing out:
1. Expertise and Workforce Training
Quantum computing is rooted in quantum physics while HPC is grounded in computer science and engineering. Bridging this cultural and technical gap requires deliberate investment in training. HPC users and support teams will need to understand fault-tolerant algorithms, quantum error correction and hybrid programming models, the report says.
A practical step is identified in dedicating classical resources to simulating eFTQC devices, complete with noise models, so that users can prototype hybrid workflows and develop intuition. Over time, maturing SDKs and compilers will abstract away many details, but early training is essential to avoid a skills gap, the report says.
2. Benchmarking and vendor evaluation
HPC centers are urged to engage actively with quantum vendors to evaluate performance metrics beyond raw qubit counts. Effective clock rate after error correction, infrastructure requirements such as power and cooling and physical footprint vary dramatically across vendors. Benchmarking will inform procurement decisions and shape vendor roadmaps, the report says.
3. Cloud and Prototyping
Cloud-based access to current QPUs offers a low-barrier entry point for HPC centers. It provides hands-on exposure to different technologies, facilitates workforce training and allows performance comparisons. However, limitations such as long queue times and restricted low-level access mean cloud use is only a steppingstone.
The report recommends deploying on-site prototypes, even with only a handful of logical qubits. Doing so develops operational expertise in cryogenics, vibration isolation and error correction — skills that will be indispensable when full eFTQC systems arrive. Prototypes also provide unique testbeds for co-designing workflows and shaping standards.
Building the Quantum-HPC bridge
For HPC centers, the transition to eFTQC requires rethinking workflows and software stacks. Unlike GPUs or FPGAs, quantum processors will not serve as drop-in replacements. Instead, they will target specific bottlenecks in broader application chains.
To unlock value, the report says HPC workflows must be re-architected to embed quantum solvers while minimizing disruption. For example, in quantum chemistry, classical methods can handle molecular structure exploration and kinetic modeling while quantum processors tackle ground-state energy calculations.
At the software level, established tools like MPI and Slurm must evolve to schedule hybrid jobs spanning CPUs, GPUs and QPUs. The report says flexibility is essential to accommodate rapid advances in both classical and quantum hardware while avoiding lock-in to proprietary toolchains that could stifle innovation.
Another challenge identified in the report lies in compilation. Today, fault-tolerant quantum algorithms are hand-optimised to map onto specific qubit layouts. Future compilers will need to automate these processes, drawing on lessons from decades of classical compiler development. This area, too, offers opportunities for HPC centers to lead, the report says.
Toward eFTQC-accelerated HPC centers
The readiness spectrum for HPC centers spans from basic awareness to full leadership. “Leaders” will deploy prototypes, co-design workflows and benchmark vendors, positioning themselves to secure scarce early access to eFTQC resources. “Fast followers” will engage later, while “late adopters” risk being left behind.
The stakes are high. Demand for early eFTQC machines is expected to far exceed supply, making strategic alliances and early engagement vital. Centers that move now will influence system design, attract top users and capture the first wave of discoveries beyond classical reach, the report says.
Conclusion: Building bridges to the next era
The arrival of eFTQC by the end of this decade will mark a transformative milestone in scientific computing, the report says.
Applications in materials science, chemistry and fusion energy are already poised to benefit, representing nearly half of workloads at leading HPC facilities. Unlike speculative scenarios, these are concrete, defined use cases embedded in existing workflows.
To seize this opportunity, HPC and quantum communities must work hand in hand. Co-developing hybrid workflows, aligning software and hardware infrastructure and training a quantum-ready workforce are essential steps. Those who prepare today will not only accelerate discovery but will shape the standards, interfaces and best practices of tomorrow.
“Now is the time to build the bridges that will carry us into the next computing era,” the report concludes. The message is clear: the future of HPC is hybrid and eFTQC is the next great accelerator.
