Unbounded Holevo additivity gaps in finite dimensions

In the world of quantum communication, information is carried by particles that can exist in multiple states at once. To send a message, a sender prepares these particles in specific ways and sends them through a channel to a receiver. A fundamental question in this field has long been whether sending two messages at the same time, using a special kind of connection between the particles called entanglement, allows for a significantly higher rate of information transfer than sending them one after another. For decades, scientists knew that this entanglement could sometimes help, but the benefit was thought to be small and limited. The prevailing belief was that no matter how large the system became, the extra information gained by using entanglement would eventually hit a ceiling. This limit was a matter of deep theoretical importance because it defined the ultimate capacity of quantum channels to carry classical data.

A new study by Jinzhao Wang has overturned this expectation by demonstrating that the advantage of using entanglement is not only real but can grow without bound. The researchers constructed a specific type of quantum channel where the benefit of sending two messages together, rather than separately, increases linearly as the system gets larger. In simpler terms, the more resources they added to the system, the more the entanglement boosted the communication speed, with no sign of stopping. This finding proves that the gap between the best possible communication rate using entanglement and the rate without it can be made arbitrarily large, even within systems that have a finite size.

The core of this discovery lies in how the researchers designed the channel. They built a system that acts like a complex filter for information. To understand the mechanism, imagine a process where a sender chooses from a large set of different operations, each one scrambling the information in a unique way. The channel then randomly selects one of these operations to apply to the incoming data. The receiver, who does not know which operation was chosen, sees a mixture of scrambled signals. The challenge is to figure out how much information can be recovered from this mixture. The researchers found that by carefully arranging the set of operations and using a specific type of quantum connection between two uses of the channel, the amount of recoverable information jumps dramatically when the two uses are treated as a single, entangled event.

To achieve this, the team relied on a mathematical framework involving random matrices and free probability, which are tools used to understand the behavior of very large, complex systems. They started by creating a theoretical model in an infinite-dimensional space, where the rules of quantum mechanics allow for a certain kind of independence between different parts of the system. In this model, they showed that the entropy, or the measure of disorder, of the output could be kept very low when the inputs were entangled, while remaining high when they were not. This difference in disorder is the key to the communication advantage. The researchers then faced the difficult task of translating this infinite model into a real, finite system that could actually be built or simulated.

They succeeded by proving that the behavior of the infinite model could be closely approximated by a finite system of a specific size. The size of this system grows in a predictable way: the number of input bits required grows with the square of the number of channel uses, while the output size grows linearly. This means that for a sufficiently large system, the researchers could construct a channel where the single-use communication rate is very small, almost negligible, yet the two-use rate becomes enormous. In fact, they showed that it is possible to create a sequence of channels where the single-use capacity shrinks toward zero, while the capacity when using the channel twice grows toward infinity. This result is striking because it shows that the ability to communicate is not a fixed property of the channel itself, but depends entirely on how the channel is used and how the signals are prepared.

The study provides a concrete recipe for building these channels, specifying exactly how many unitary operations, which are the quantum equivalent of rotations, are needed to create the effect. The researchers used a probabilistic method to show that such channels exist, meaning that if one were to randomly select the necessary operations from a large pool, the chance of finding a set that works is extremely high. While they have not yet provided a fast algorithm to find these specific operations in practice, the proof of their existence is mathematically rigorous. The work also establishes that the gap in communication rates scales linearly with the number of output bits, a result that was previously unknown.

This breakthrough changes the understanding of quantum communication limits. It confirms that entanglement across multiple uses of a channel is not just a minor correction but a powerful resource that can be scaled up indefinitely. The findings suggest that the classical capacity of a quantum channel, which is the maximum rate at which information can be sent, is much more complex than previously thought. It depends on the ability to coordinate signals across multiple uses in a way that exploits the unique properties of quantum mechanics. The researchers' work closes a long-standing question about whether this advantage is bounded, showing instead that it is unbounded.

The implications of this work are primarily theoretical, reshaping the landscape of quantum information theory. It demonstrates that the rules governing how information flows through quantum systems are more flexible and powerful than earlier models suggested. By showing that the gap between single-use and multi-use capacities can be made as large as desired, the study opens new avenues for exploring the fundamental limits of communication. It also highlights the importance of structured randomness in quantum systems, where carefully designed disorder can lead to highly ordered and efficient information transfer. The results stand as a definitive proof that in the quantum realm, the whole can be vastly greater than the sum of its parts, and that this difference can grow without limit.