The Signal
UCF Wins $400,000 DOE Grant for Partially Fault-Tolerant Tools
UCF assistant professor Siyuan Niu’s team received a $400,000 Department of Energy grant to develop software for partially fault-tolerant quantum computers. The planned open-source framework will compile electron simulations, decode errors
Quantum.world Team · Written with AI · fact-checked ·
“For this project, the task is simulating how electrons behave in materials.”
Siyuan Niu, assistant professor in UCF’s Department of Electrical and Computer Engineering
A $400,000 project starts with electron simulation
The University of Central Florida’s Siyuan Niu is leading a project to build software tools for a type of quantum computer that uses less hardware than a fully fault-tolerant machine. The Department of Energy is supporting the work with a $400,000 grant, UCF’s College of Engineering and Computer Science announced on October 20. UCF co-investigators Di Wu and Eduardo Mucciolo are also part of Niu’s team.
The project asks whether partial protection from errors can be enough to run useful simulations without the hardware and time demands of full fault tolerance. Its target is the simulation of electrons in materials, a problem the team will use to compare machine designs and estimate what a calculation would require. The goal is not to announce a quantum-computing result, but to assess whether this approach can perform the simulation accurately and quickly enough to be worthwhile.
STAR protects some operations, not all
Niu’s team will study the Space-Time Efficient Analog Rotation architecture, known as STAR. The design applies full quantum error correction to basic operations, while performing heavily used rotation operations directly as analog rotations with only partial protection. Those rotations are central to the simulations the project targets.
The compromise could reduce the extra qubits and time needed to perform rotations in fully fault-tolerant designs. But the analog rotations retain some error and may need to be repeated until they succeed. That means the software must account for how those operations are managed and scheduled; UCF says the compiler and cost estimates will be tailored to that trade-off.
The Fermi-Hubbard model supplies the benchmark
The team will use the Fermi-Hubbard model, a standard framework for studying interacting electrons, to test whether a partially fault-tolerant computer can simulate how electrons evolve over time. Error correction will remain part of the test: the computer must identify and correct faults even though individual qubits and operations sometimes fail.
In the approach described by UCF, information is spread across multiple physical qubits, checks are repeated to detect errors, and a classical decoder identifies corrections. The decoder matters because useful simulations require many operations, giving errors repeated opportunities to accumulate. Niu’s project will examine how quickly and accurately the process can identify errors alongside the computing resources it consumes.
Three software components will estimate the cost
By the project’s end, the team aims to release an open-source framework with three parts: a quantum simulation compiler, a GPU-based decoder and a resource estimator. The compiler translates a simulation into quantum operations; the decoder processes error information; and the estimator will calculate the qubits, time, classical computing power and operations required for an electron simulation.
UCF says the framework will be modular, allowing other researchers to plug in their own compilation or decoding methods. That design would let researchers compare different software choices within the same assessment, rather than treating one compiler or decoder as fixed. The announcement does not give a delivery date or report performance results from the tools.
Niu calls the project an assessment, not a victory claim
The framework is intended to compare STAR-based simulations with today’s noisy, unprotected devices and fully protected designs. Those comparisons could help researchers judge which simulations are realistic and which hardware improvements matter most, according to Niu. The work also addresses the broader question of quantum utility: whether a machine can deliver scientifically useful calculations at a practical cost.
Niu describes the intended outcome as “a roadmap and assessment,” not a guaranteed demonstration of quantum advantage. The distinction sets the project’s measure of success: whether its compiler, decoder and estimator can make the requirements of a specific materials simulation clearer. For now, UCF’s announced deliverable is an open-source software framework built to quantify those requirements.