⚛️ quantum physics

A 12-CNOT Double Qubit Excitation Gate

This paper presents the first 12-CNOT decomposition of the double qubit excitation operator, achieving state-of-the-art performance by minimizing CNOT count, CNOT depth, and total circuit depth while maintaining a near-optimal one-qubit gate count.

Original authors: Irfansha Shaik

Published 2026-08-13

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Irfansha Shaik

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where computers don't just crunch numbers but dance with the very fabric of reality. This is the realm of quantum computing, a field that promises to solve problems too complex for today's supercomputers. But here's the catch: these quantum machines are incredibly fragile. To make them work, scientists have to build "gates," which are like the switches and levers of a quantum circuit. Think of these gates as the instructions in a recipe; if you use too many steps or the wrong tools, the delicate ingredients (the quantum information) get ruined before the dish is ready. One of the most important, yet tricky, ingredients in this recipe is something called a "double qubit excitation." It's a specific move that swaps energy between two pairs of particles, a crucial step for simulating how molecules behave, which could help us discover new medicines or better batteries. The challenge has always been how to perform this swap using the fewest possible "CNOT" gates—a type of quantum instruction that is notoriously difficult to execute perfectly. The fewer CNOTs you use, the less likely the computer is to make a mistake, making the whole process faster and more reliable.

In this paper, a researcher named Irfansha Shaik from Copenhagen presents a clever new way to perform this tricky double qubit swap. For a long time, the best-known method to build this gate required 13 of those difficult CNOT instructions. It was like trying to cross a river using 13 stepping stones; you could get across, but it was risky and slow. Shaik has now discovered a new circuit design that does the exact same job using only 12 CNOTs. To put it in perspective, imagine finding a shortcut that saves you one whole step in a long, winding maze. This new design doesn't just save a single step; it also shortens the time the computer needs to think about the problem (the "depth") and keeps the total number of steps in the entire recipe lower than any previous attempt. While the author notes that this is the first time such a 12-CNOT version has been reported, they also point out that it only adds two extra simple "one-qubit" moves compared to the most efficient previous designs, making it a very balanced and practical improvement.

The paper walks us through how they got there. First, they looked at the "standard" way to do this, which would have taken a whopping 48 CNOTs—like trying to cross that river by building a bridge out of 48 stones, which is way too heavy and slow. Then, they examined the "state-of-the-art" (the current best) methods, which had already been trimmed down to 13 CNOTs. These previous methods were like different teams of engineers trying to optimize the bridge, but they all hit a wall at 13 stones. Shaik used various digital tools and mathematical tricks to explore different ways to arrange the circuit. They found a new arrangement that shaves off that final, stubborn CNOT.

The result is a circuit that is not only shorter in terms of the number of CNOTs (12 instead of 13) but also faster to run. The new design has a "CNOT depth" of 10, meaning the longest chain of these difficult steps is shorter than in previous versions. The total length of the entire circuit is just 16 steps, which is the lowest recorded so far. While the paper doesn't claim this is the absolute final answer to the universe's problems, it presents this 12-CNOT circuit as a significant, verified improvement over everything that came before it. It's a small but vital step forward, proving that with the right creative approach, we can make quantum computers a little bit more efficient and a little bit less prone to errors.

Technical Summary: A 12-CNOT Double Qubit Excitation Gate

Problem Statement
The practical implementation of high-level quantum gates is a critical bottleneck for both near-term and fault-tolerant quantum algorithms. Specifically, the Double Qubit Excitation Operator ( $U_{klij}(\theta)$ ) serves as a fundamental building block for variational ground-state ansätze (e.g., UCCSD in VQE, QEB-ADAPT-VQE, FAST-VQE) and fault-tolerant applications like Trotterized Hamiltonian simulation and quantum phase estimation (QPE). This operator implements a continuous rotation between the computational basis states $|0011\rangle$ and $|1100\rangle$ within a 4-qubit system.

The challenge lies in decomposing this operator into native gate sets (specifically CNOT and single-qubit gates) with minimal resource overhead. Prior to this work, the state-of-the-art (SOTA) implementations required 13 CNOTs, with trade-offs existing between CNOT count, CNOT depth, total circuit depth, and single-qubit gate counts.

Methodology
The paper approaches the decomposition problem by analyzing the structure of the double excitation operator, which can be viewed as a triple-controlled $R_y(2\theta)$ rotation sandwiched between two CNOT circuits ( $L$ and $R$ ) that map the target states $|0011\rangle$ and $|1100\rangle$ to $|0111\rangle$ and $|1111\rangle$ , respectively.

Baseline Construction: The author first establishes a baseline using a "Gray-code expansion" of the $C^3R_y(2\theta)$ rotation. This standard approach yields a 14-CNOT circuit.
Literature Review: The author reviews existing SOTA 13-CNOT decompositions (e.g., by Yordanov et al., Nam, and Wang), noting that these typically involve rewriting rules to absorb CNOTs from the $L$ and $R$ circuits into the rotation decomposition.
Optimization and Synthesis: To achieve a lower CNOT count, the author employed various circuit synthesis and optimization tools, including Q-Synth, Qiskit Transpiler, and Tket. They utilized advanced synthesis techniques such as Clifford Synthesis, CNOT+ $R_z$ Synthesis, and KAK decomposition to explore the solution space and eliminate suboptimal paths.

Key Contributions and Results
The primary contribution of this work is the presentation of the first reported 12-CNOT decomposition of the double qubit excitation operator. The proposed circuit (illustrated in Figure 5 of the paper) achieves the following metrics:

CNOT Count: 12 (improving upon the previous SOTA of 13).
CNOT Depth: 10 (improving upon the previous SOTA of 11).
Total Circuit Depth: 16 (improving upon the previous SOTA of 20–22).
Single-Qubit Gate Count: 13. While this is slightly higher than the lowest previous count (11 by Nam), the author notes this is only a marginal increase of 2 gates to achieve significant reductions in CNOT resources.

The paper provides a comparative analysis in Table 1, demonstrating that the new circuit outperforms all previous SOTA implementations across the three primary metrics of CNOT count, CNOT depth, and total depth.

Significance and Claims
The author claims that this work represents a significant step forward in the efficient implementation of quantum algorithms. By reducing the CNOT count and depth, the new decomposition directly addresses the noise and error accumulation issues inherent in current quantum hardware. The paper modestly asserts that this is the "first reported" implementation with fewer than 13 CNOTs, offering a more resource-efficient alternative for both variational algorithms and fault-tolerant simulations without introducing prohibitive overhead in single-qubit gates. The work relies on established synthesis tools and mathematical decompositions rather than proposing new hardware or experimental protocols.

Technical Summary: A 12-CNOT Double Qubit Excitation Gate

More like this