January 28, 2026 -- The capabilities of quantum computing are likely to increase dramatically in the next five to ten years, leading to significant gains in certain scientific areas, says a new landmark report from quantum computing researchers at the National Energy Research Scientific Computing Center (NERSC), a U.S. Department of Energy (DOE) user facility located at Lawrence Berkeley National Laboratory (Berkeley Lab).

NERSC serves over 12,000 science researchers under the umbrella of the DOE Office of Science (SC), offering high performance computing (HPC) resources for a broad range of scientific applications. Quantum computing is an up-and-coming computing technology with the potential to solve specific types of scientific problems, some of which are considered unsolvable with current technology. It’s likely to play a role in future HPC systems – although exactly what role, and when, remains to be seen. The report, “Quantum Computing Technology Roadmaps and Capability Assessment for Scientific Computing,” aims to offer some measure of clarity about potential applications and timelines for development, as well as the current state of the technology and the industry around it.

“New approaches to computing are desirable in a post-Moore’s Law world, and quantum computing may be among the most interesting alternative approaches,” said NERSC quantum computing engineer Daan Camps, an author on the paper. “Future NERSC systems may have quantum computing components, and it’s essential for us to understand how they might serve the SC workload and what the roadmap for performant quantum computers might be.”

To get an idea of which research domains might benefit most from quantum computing, the researchers analyzed the current workload at NERSC and found that over 50% of jobs performed on Perlmutter, NERSC’s current flagship system, fall into the categories of materials science, quantum chemistry, and high-energy physics. All of these disciplines are likely to see major improvements in solving speed, complexity of solvable problems, and accuracy of solutions with the growth of quantum computing technology. In particular, these domains tend to deal in complex many-body problems, which are difficult to solve with classical methods but conducive to quantum computing. The researchers say it’s unclear whether quantum computing will be applied to the broad population of SC users or more applicable to a smaller subset of very difficult problems, but their findings show that performant quantum computing might provide significant benefits to the SC workload once the technology is mature.

Surveying the landscape

In the report, the team gathered 140 resource estimates for benchmark problems from the scientific literature and found three things. First, Hamiltonian simulation, a quantum computing method that measures the energy of a quantum system over time, is the linchpin of many quantum computing processes and is likely to be crucial to progress. They also found that the amount of compute resources required for success varies dramatically between scientific domains, with high-energy physics being the most resource-intensive of the most common quantum domains. However, they also found that algorithmic improvements over the last five years have already significantly reduced the compute resources required. If that trend continues as expected, applications across all domains are likely to advance rapidly.

To get an idea of the quantum landscape outside DOE SC, the researchers also surveyed public technology roadmaps from ten different quantum vendors. All predicted an exponential increase in the capabilities of quantum computers over the coming decade, up to nine orders of magnitude – potentially a good match to the compute resource needs identified by the researchers in their survey of benchmark problems. Although roadmaps and timelines aren’t exact, they say expected improvements in hardware and algorithms in the coming years will almost certainly unlock new capabilities at NERSC.

Finally, the researchers studied the possible timescales for a range of jobs running on a variety of quantum technologies and proposed a model, Sustained Quantum System Performance (SQSP), to compare possible system throughput across different systems, technologies, and benchmark workloads. The model may be a useful tool in helping those in the quantum community to make decisions regarding their own paths to quantum computing.

Moving forward

Overall, the study indicated that although developing applications for quantum computers remains challenging, the technology is progressing and is likely to significantly benefit some portion of the SC workload in the coming years, especially in certain scientific domains.

“Solving quantum many-body problems is a large part of the NERSC workload, so quantum computing is of great interest,” said Camps. “Developing applications for quantum computing remains a challenge, but improvements to hardware and algorithms are already occurring, and if that trajectory continues, we believe it will open up new possibilities for NERSC users.”

Among the next steps, according to NERSC quantum computing researcher Ermal Rrapaj, a co-author on the paper, is improving the landscape of quantum algorithms to allow for widespread use.

“In the next five to 10 years we see a strong need for continued algorithmic research and development to reduce quantum resources and allow for wider scientific impact given the expected hardware capabilities from publicly available roadmaps,” said Rrapaj. “Only time will tell.”