Search

Bridging the Quantum Gap

Art courtesy of Tai Michaels

Art courtesy of Tai Michaels.

Quantum computers may one day solve difficult problems, but currently they still face many hardware limitations. Quantum physics—the science of the very small, such as electrons and photons— allows for “qubits” that can represent “0” and “1” at the same time, a concept known as superposition. This gives them exponentially faster processing speeds than traditional computers, which use binary electrical bits assigned a fixed value of either “0” or “1.”

However, today’s quantum processors can only work with a few qubits, compared to the billions of transistor bits in traditional computers. To increase the number of qubits involved, researchers from the Max Planck Institute of Quantum Optics in Germany have demonstrated a distributed quantum logic gate that can connect multiple quantum processors. Severin Daiss, Gerhard Rempe, and colleagues set up two qubits, located sixty meters apart in two labs, to interact using an additional photon sent between them. Quantum logic gates like this are the quantum version of classical logic gates, performing specific operations to input qubits. Researchers hope having a distributed setup—where the quantum logic gate is not necessarily in one location only—provides a more flexible, modular approach towards achieving larger quantum computing power. 

Using the distributed gate, the German researchers produced pairs of maximally entangled qubits known as Bell states. Entangled qubits can be highly correlated even though they are physically apart. Such “spooky action at a distance,” as Einstein remarked, actually provides quantum computers with their potential power. If one element in a pair of Bell state qubits is measured, the other qubit’s condition is also fixed, no matter how far away. 

In an ordinary computer, billions of transistors work in unison to perform logical operations—such as addition, multiplication and information processing—on electrical bits. Quantum researchers have adopted a similar approach to quantum computing (among other methods) to design “quantum circuits.” Quantum circuits manipulate qubits, which, in contrast to electrical bits, are not a fixed “0” or “1.” The circuits therefore require quantum logic gates instead of the silicon chips found in ordinary computers. 

As an example of a nonlocal operation, Daiss and Rempe performed a quantum controlled-NOT (CNOT) gate in their experiments. The CNOT gate uses the input of one qubit to determine whether or not to invert the input of a second qubit. In many quantum circuits, CNOT gates are used to achieve the superposition necessary for any potential quantum advantage over classical computing. 

Daiss and his colleagues placed two optical cavities containing a rubidium atom in two separate rooms within a building at the Max Planck Institute. In order to perform a quantum logic gate between the two devices, the researchers first launched a photon towards the first optical cavity along an optical fiber, then directed the reflection to the second cavity (in the other room). The reflection caused an interaction between the photon state and each atomic qubit state. Following the photon measurement, with the addition of a Z-gate—a type of quantum logic gate—on the first cavity controlled by a classical communication channel, Daiss and his team completed a CNOT gate. 

“Our experiment is the first that does such a gate between qubits residing in completely independent laboratories,” Daiss said. 

The team reported results of up to eighty-five percent accuracy for the CNOT operation, at a computation rate of one kHz. While this is low compared to current state-of-the-art quantum hardware, the approach holds promise for a modular approach to scaling up quantum computers. Qubits are much more fragile than electrical bits; the slightest measurement by environmental disturbances (dust, air, stray light) could break the trance. As qubits typically reside in dilution fridges or vacuum chambers—which have limited space—increasing qubit number requires denser packing. This can cause problems like crosstalk noise and limited access to each individual qubit, complicating manipulation and measurement. Thus, in the future, instead of just increasing just the number of qubits in a single processor, researchers may connect multiple quantum devices using distributed quantum gates to improve computational power. “Such a modular approach might open a new development path for larger quantum computers,” Daiss said. 

The researchers believe that their protocol with the use of a separate photon may be applied to other quantum operations and gates between distant qubit modules. In a future quantum computer, it is likely that many quantum gates—such as the CNOT—will be involved, each with its own chance of failure. Thus, the photon’s ultimate arrival would signal a successful quantum operation, before subsequent operations proceed.