Two New Simulators Tease Future of Quantum Computing

A 53-qubit and 51-qubit quantum simulator point to larger arrays on the path to universal quantum computing

An artist's depiction of a quantum simulation. Lasers manipulate an array of more than 50 atomic qubits to study the dynamics of quantum magnetism.
Illustration: E. Edwards/JQI
Advertisement

A universal quantum computer capable of outperforming today’s classical computers in solving many different problems remains the biggest future prize for many engineers and researchers. One possible path toward that goal comes from two U.S. research groups that have demonstrated some of the largest quantum simulators ever built. Such specialized devices are much less versatile than the vision for universal quantum computers, but share architectural similarities that could pave the way for the latter.

Quantum simulators are designed to tackle very specific problems in scientific fields such as high-energy physics and chemistry. These devices have mostly consisted of small arrays of five or 10 quantum bits (qubits) that can each represent multiple states of information simultaneously. In recent work, one research group used lasers as optical tweezers to assemble a 51-qubit array of so-called Rydberg atoms. A second group showed how to build a 53-qubit “trapped ion” device using electric fields to control a string of charged atoms. 

“This is a very exciting time in the development of quantum technologies,” says Ahmed Omran, a postdoctoral quantum physics researcher at Harvard University and coauthor on the Rydberg atom paper. “Our work, along with that of [the University of Maryland and the National Institute of Standards and Technology], and that being done elsewhere—including other universities, as well as private companies, such as Google, IBM, Microsoft—is taking a step forward towards using the fine control of a few particles and taking it into a realm where we can scale it to system sizes that might show a quantum speedup over classical devices in solving important problems.”

The two independent research groups took the first steps in demonstrating some basic experiments with a large numbers of controlled qubits. Omran and his colleagues from Harvard, MIT and the California Institute of Technology detailed their 51-qubit “Rydberg atom” device in the 29 November 2017 online issue of the journal Nature. That same issue includes a second paper on the 53-qubit “trapped ion” device built by the team from the University of Maryland and the National Institute of Standards and Technology.

To build their 51-qubit device, Omran and his colleagues used 101 lasers as optical tweezers to shine their light upon a “dilute vapor of rubidium atoms.” Each laser beam had a 60 percent chance of trapping a single atom, which meant the initial array of atoms looked very random. One of the group’s key achievements was coming up with a way to rearrange the laser tweezers on the fly so that they could “create perfect atom arrays of any desired size and pattern with up to 51 particles.”

The next stage involved turning the atoms into qubits. Researchers accomplished this by focusing additional lasers on individual electrons tightly orbiting the atomic nucleus. That provided the energy boost necessary to push the electron out to a much larger orbit—a Rydberg state—without ripping the electron away from the atom entirely. The interactions between the Rydberg atoms then allowed researchers to effectively manipulate them as qubits. 

The distance between atoms determines the interaction strength, and since we can control the position of each atom individually, we can program various interaction patterns and study the evolution of this quantum many-body system,” Omran explains.

The “trapped ion” quantum simulator demonstrated by the second research group took a somewhat different approach to building a 53-qubit array. In that case, researchers created a single-file line of charged atoms (ions) of ytterbium. Such ions typically want to repel each other because they have the same charge. But the research group used electric fields to counteract that repulsion tendency.

Pioneering researchers such as Christopher Monroe (a co-author on the trapped ion paper) have helped create a “great toolbox and understanding” for controlling qubit arrays based on the trapped ion approach, according to Omran and his colleagues. They added that trapped ion arrays also enable researchers to “widely tune the range of their interactions.”

But a main challenge for trapped ion arrays involves offsetting the strong repulsion interactions between individual ions—something that can complicate efforts to build ever-larger quantum simulators.

By comparison, the Rydberg atom approach presents more unknowns in terms of how Rydberg states behave and can be controlled. But its advantage comes from using neutral atoms that can be packed together in large numbers before switching on the Rydberg state interactions. “While we currently are at a similar system size to the largest trapped ion system, there are some very promising near-term routes that will allow us to do these types of experiments with hundreds of atoms,” Omran says.

In both cases, the research groups have taken huge strides in building the larger arrays of atomic qubits and beginning to investigate their scientific usefulness. But there is much more work to be done before researchers can show they have mastered control over the quantum interactions among these 50-qubit arrays.

A next step for Omran and his colleagues involves studying quantum entanglement—the “spooky action at a distance” phenomenon that allows two entangled particles to remain perfectly linked despite physical separation—because it can naturally arise from the interactions in their Rydberg atom array. Quantum entanglement is generally considered crucial for future quantum computers to perform many different calculations simultaneously.

The researchers also want to improve their control over individual atoms as qubits. One of the biggest long-term goals for such quantum systems involves improving coherence times, which in this case means how long atoms can function as qubits by maintaining certain states. “As we learn to better control our systems—both Rydberg simulators and ion simulators—we will be able to improve our coherence properties and perform more sophisticated simulations,” Omran says. (Omran’s co-authors, including Hannes Bernien, Alexander Keesling and Harry Levine, also contributed detailed responses to questions posed by IEEE Spectrum.)

Tech giants such as Google and IBM have been pursuing a completely different approach to quantum computing systems by building superconducting qubits upon integrated circuits. Google hopes to show that its 49-qubit system can achieve quantum supremacy by proving quantum computing can indeed solve problems that would be impractical for classical computing. By comparison, IBM engineers believe that achieving quantum supremacy will require at least 57 qubits or more.

But even that long-awaited demonstration of quantum supremacy would just be an early step on the road toward universal quantum computing. Researchers generally believe that they will need very large arrays of qubitstens of thousands or even millionsin order to make universal quantum computing a more practical tool. Until that time, the smaller arrays of qubits designed to solve more specific quantum simulation problems could still prove quite handy.

The Tech Alert Newsletter

Receive latest technology science and technology news & analysis from IEEE Spectrum every Thursday.

About the Tech Talk blog

IEEE Spectrum’s general technology blog, featuring news, analysis, and opinions about engineering, consumer electronics, and technology and society, from the editorial staff and freelance contributors.