Abstract
Given a quantum gate $U$ acting on a bipartite quantum system, its maximum (average, minimum) entangling power is the maximum (average, minimum) entanglement generation with respect to certain entanglement measure when the inputs are restricted to be product states. In this paper, we mainly focus on the 'weakest' one, i.e., the minimum entangling power, among all these entangling powers. We show that, by choosing von Neumann entropy of reduced density operator or Schmidt rank as entanglement measure, even the 'weakest' entangling power is generically very close to its maximal possible entanglement generation. In other words, maximum, average and minimum entangling powers are generically close. We then study minimum entangling power with respect to other Lipschitiz-continuous entanglement measures and generalize our results to multipartite quantum systems. As a straightforward application, a random quantum gate will almost surely be an intrinsically fault-tolerant entangling device that will always transform every low-entangled state to near-maximally entangled state.
Introduction
In quantum information theory, entanglement serves as the fundamental "fuel" for critical tasks such as quantum teleportation, superdense coding, and quantum cryptography. The efficiency of generating this resource is quantified by a gate's entangling power. Traditionally, researchers have focused on the maximum or average entangling power of a unitary operator. However, these metrics can be misleading; a gate might perform exceptionally well on average but possess "blind spots"—specific input product states for which it fails to generate any entanglement.
This paper addresses the "weakest link" in quantum gate performance: the minimum entangling power. By investigating high-dimensional bipartite and multipartite systems, we demonstrate that for a random quantum gate, the gap between its worst-case and best-case performance effectively vanishes.
Quantifying Entangling Power
To understand the capabilities of a quantum gate $U$, we categorize its entangling power into three distinct metrics based on the entanglement generated from input product states:
- Maximum Entangling Power: The highest possible entanglement produced from a specific pair of product states.
- Average Entangling Power: The entanglement produced averaged over the entire manifold of possible input product states.
- Minimum Entangling Power ($P_{min}$): The smallest amount of entanglement the gate produces, representing its guaranteed performance regardless of the input state.
If $P_{min}$ is zero, the gate is not a reliable entangler because it cannot guarantee entanglement for all inputs. This research explores whether $P_{min}$ is generically close to the maximum possible value in high-dimensional Hilbert spaces.
Mathematical Framework
The study employs a diverse set of mathematical tools to bound the probability of encountering a "bad" gate (one with low minimum entangling power):
- Concentration of Measure and Riemann Geometry: Using the Ricci curvature of the special unitary group $SU(N)$ and log-Sobolev inequalities, the authors show that the entangling power of a random unitary is tightly concentrated around its mean. This suggests that deviations from high entangling performance are exponentially rare in high dimensions.
- Algebraic Geometry: For discrete measures like the Schmidt rank, the authors model the set of product states as a "Segre variety." By calculating the dimensions of these varieties, they prove that a random rotation in Hilbert space will almost certainly map the entire set of product states away from low-entanglement regions (determinantal varieties).
- Multipartite Generalization: The results are extended to systems with $N \geq 3$ subsystems using tensor rank and secant varieties, showing that random gates produce near-maximum entanglement across any bipartite cut.
Key Insights: The Vanishing Gap
The most significant finding is that for large quantum systems, $P_{min} \approx P_{avg} \approx P_{max}$. In high-dimensional spaces, the "volume" of states with low entanglement is so infinitesimally small that a random unitary transformation is statistically forced to move all product states into the high-entanglement region.
Universal and Fault-Tolerant Entanglers
This leads to the concept of a Universal Entangling Device. In high dimensions, one does not need to carefully calibrate the input state; almost any product state fed into a random gate will emerge nearly maximally entangled.
Furthermore, these gates exhibit intrinsic fault tolerance. Because they entangle all states effectively, they are robust against errors in state preparation. Even if an input state is slightly "off" (e.g., a low-entangled state instead of a pure product state), the gate will still reliably produce a near-maximally entangled output.
Results by Measure
- Von Neumann Entropy: For a bipartite system $d_A \otimes d_B$, the minimum entangling power is shown to be at least $\log d_A - \frac{d_A}{d_B \ln 2} - 1$ with high probability as dimensions increase.
- Schmidt Rank: The authors provide an exact formula (Theorem 26) for the generic minimum Schmidt rank, proving that a random gate will almost surely reach a specific, high-rank threshold for all inputs.
Limitations and Future Directions
While the existence of these "universal entanglers" is mathematically guaranteed for random gates, two primary challenges remain:
- Verification Complexity: Determining if a specific given gate has a non-zero minimum entangling power is computationally difficult. It requires calculating a Gröbner basis, a process that can scale exponentially with the system's dimension.
- Small Systems: The concentration of measure effects only become dominant as the Hilbert space dimension grows. For small systems, such as two-qubit gates, the minimum entangling power can still be zero.
- Constructive Recipes: The current work relies on a "random-pick" strategy. Developing deterministic, simple circuit architectures that achieve these bounds without relying on randomness remains an open challenge for the field.
Summary updated