Applications
Halving the Cost of Quantum Applications, Again: Xanadu's Latest QROM Breakthrough
By Danial Motlagh · October 20, 2026
Xanadu’s May 2026 Quantum Read-Only Memory (QROM) breakthrough lowered the resource cost of a key building block for quantum applications by roughly half. Now, the Xanadu algorithms team has halved it again.
by Danial Motlagh
Data loading, transferring classical information onto quantum hardware, is a critical component of virtually every quantum algorithm. Historically, this step has presented a major cost and hardware bottleneck, requiring significant quantum computational resources. QROM is an indispensable subroutine, serving as the primary gateway for data loading, and despite its critical role across quantum algorithms, QROM performance had reached a plateau for nearly seven years.
Earlier this year, Xanadu announced an algorithmic breakthrough in QROM, breaking this multi-year plateau and introducing fundamental optimizations to data loading. Specifically, this work reduced the number of Toffoli gates, which are among the most computationally expensive operations a quantum computer can perform. The advancements Xanadu made on QROM thus far offered immediate benefits for near-term utility-scale quantum computers (USQC), where making use of the limited number of available qubits is crucial to enabling industry use cases.
Building upon those foundational optimizations, new research published by the Xanadu algorithms team on the arXiv introduces dense encoding, a novel technique that pushes QROM efficiency even further.
Key Highlights
- 4x Cumulative Cost Reduction: Xanadu’s combined algorithmic innovations achieve up to a 4-fold reduction in compute requirements compared to legacy industry approaches.
- More Impactful Problems on Smaller Hardware: By making better use of the limited number of available qubits, we believe dense encoding will address one of the primary bottlenecks facing near-term hardware, and we expect it will unlock more impactful commercial use cases with the same machine.
- Establishing a New Benchmark for QROM Efficiency: Halves the computational cost of the previous state-of-the-art benchmarks established by Xanadu.
By increasing algorithmic efficiency, Xanadu is actively reducing the size and cost of hardware required to run commercial quantum applications.
Fault-tolerance refers to a quantum computer’s ability to carry out a computation correctly despite the presence of noise (errors). Quantum systems’ delicate nature leaves them more prone to noise than classical computers, making quantum error correction and fault-tolerance a necessity for any useful computation. Within the fault-tolerant paradigm, a certain type of operations referred to as “non-Clifford gates”, become the main bottleneck. Toffoli gates are the prominent example of such operations.
It turns out that in almost all practical applications of quantum computers, the dominant source of these expensive gates comes from data loading or QROM. Over the years, people have found ways to reduce the cost of QROM by making use of “dirty qubits”, qubits that are already in use for a different part of the computation and thus don’t introduce any additional overheads. Both of our recent breakthroughs are based on clever utilizations of such dirty qubits to achieve an even larger reduction in the cost of QROM under the same dirty qubit budget, but are based on different insights.
Our earlier "sequential bit packets" result achieves this by breaking up a larger lookup table into a sequence of smaller ones to make better use of dirty qubits by sharing table passes across the smaller tables. This provided a 2x speed up over previous methods. Our recent innovation makes better use of dirty qubits in a new way: storing two bits of classical information in each dirty qubit instead of one. It does this by utilizing both the binary value of the dirty qubit, as well as its “phase”. We call this idea dense encoding. On its own, dense encoding offers another 2x speed up over the previous methods and a 4x speed up when combined with our earlier sequential bit packets construction.
By halving the cost of QROM for the second time in under a year, a 4-fold reduction overall, Xanadu believes we are able to significantly lower the capital expenditure and physical infrastructure needed to scale quantum capabilities, bringing USQC one step closer for the industry at large.
Browse these recommended resources to dive deeper into our scientific advancements.
Published on October 20, 2026
