How NIST Helped Start an Industry: Our Role in Jump-Starting Quantum Information Science

Credit: Courtesy of Chris Monroe

If you love all things quantum like I do, you may also follow the latest news about quantum computers and the larger field of quantum information. Every week, it seems, researchers announce a breakthrough in this realm. Today, industry, academia, government agencies and other organizations all contribute to a thriving quantum ecosystem.

What you may not know is that our “little-known” federal science agency — NIST — played a big role in jump-starting the entire field of quantum information science in the 1990s.

So in this International Year of Quantum Science and Technology, I would like to tell you a few stories about how that happened. Along the way, I’ll share some insights directly from a few of the pioneers in the field.

Quantum Information Falls Into NIST’s Lap

In a 1994 theoretical paper, the mathematician Peter Shor found that computers that operated with the rules of quantum mechanics could potentially break the conventional encryption that safeguards electronic data from our emails to our bank accounts. This theoretical discovery showed that sufficiently powerful quantum computers, if they could be built, could have major real-world consequences.

Meanwhile, NIST researchers already had qubits — the building blocks of quantum computers — without knowing it.

They were inside atomic clocks.

A group led by then-NISTer and future Nobel laureate Dave Wineland discovered that it could apply years of work on atomic clocks to quantum computing.

“The field of quantum information sort of fell in our laps,” recalled Duke University physicist Chris Monroe, who worked with Wineland as an early-career staff scientist at NIST in the mid-1990s. “We were playing around with atoms in a certain way that would make them better for atomic clocks.”

The NIST lab where Wineland worked from 1975 to 2018 was (and still is) fertile ground for atomic clock research — and its fruits provided many of the tools needed to create the building blocks for quantum computers.

The Computer-Clock Connection

To understand how quantum computers are connected to atomic clocks, let’s talk about how computers handle information.

Computers convert information such as words, numbers and pictures into a series of 0s and 1s. These 0s and 1s are called binary digits, or bits. 

Traditional computers store bits in devices such as magnetic hard drives, consisting of tiny chunks of magnetized material known as domains. A “0” can be represented by a domain whose magnetic field is pointing down. A “1” is a domain with a magnetic field pointing up. The drive reads the data and has a “write head” that can change a 0 to a 1 or vice versa by flipping the direction of the magnetism. A traditional bit can’t be both digits at the same time. 

Qubits, or quantum bits, are made with individual atoms or other objects that follow the rules of quantum physics. Atoms can carry a lot of energy, like a hot potato, or they can have a minimum amount of energy, like an ultracold potato. An atom in a lower-energy state can represent a 0. The same atom in a higher-energy state can represent a 1. 

According to quantum physics, an atom can be in a lower- and higher-energy state at the same time, a phenomenon known as superposition. That means a quantum bit can be a 0 and 1 simultaneously.

Whereas researchers can perform only one calculation at a time with traditional bits, they can perform multiple calculations simultaneously with qubits. A quantum computer containing many qubits has the potential to do certain tasks in a fraction of the time it would take traditional computers.

As it happens, atomic clocks work by putting atoms in a superposition of low- and high-energy states — making some atoms ideal candidates to serve as quantum bits.

NIST Holds the First Workshop on Quantum Information

In the early 1990s, at the NIST laboratories in Boulder, Colorado, Wineland and his colleagues were busy working on their latest atomic clocks based on atomic ions (charged atoms). Around that time, physicists who were aware of quantum information were few and far between.

One of those exceptions was Daniel Kleppner, an atomic physicist at MIT, who had supervised the Ph.D. thesis of future NIST Nobel laureate Bill Phillips. At MIT, Kleppner heard a fascinating talk by Artur Ekert on quantum information.

Kleppner recommended that Ekert be invited to speak at the International Conference on Atomic Physics, to be held in Boulder in the summer of 1994.

“I was given a completely different congregation to preach to, but I felt I was preaching to the converted,” said Ekert, then a postdoc at the University of Oxford. “I was amazed how enthusiastic the atomic physicists were about the subject of quantum computation.”

Meanwhile, NIST physicist Charles Clark organized the NIST Workshop on Quantum Computing and Communication, taking place in Gaithersburg, Maryland, that August. As the first major workshop devoted to these topics, it’s recognized today as a landmark event.

Shortly after these meetings, theoretical physicists Ignacio Cirac, now at the Max Planck Institute of Quantum Optics in Germany, and Peter Zoller of the University of Innsbruck in Austria came up with an idea for making quantum computers a reality in the laboratory. A few years earlier, each of them had spent time at JILA, the joint institute of NIST and the University of Colorado Boulder. Inspired by Ekert’s talk, their May 1995 paper envisioned using individual ions as quantum bits for quantum computing in an experimental system similar to what the NIST atomic clock physicists had built.

When Wineland and colleagues saw this paper, they immediately saw the parallels between controlling atoms in an atomic clock and performing a quantum computation.

“The kind of technology we use is very much the same,” he said.

The First Quantum Logic Gate With Individual Qubits

Logic gates are the building blocks for processing information in computers. They take inputs of 0s and 1s, process those bits, and produce outputs, also in the form of 0s and 1s.

There are many types of logic gates, and they process data in different ways. When connected, they can perform the wide variety of tasks that computers can do, from crunching numbers in spreadsheets to creating AI chatbots that respond to your inputs.

In 1995, the NIST group made the first quantum version of a logic gate, using individual qubits in beryllium ions. They used lasers to link or “entangle” two of these qubits, so that they interacted to produce an output, just like a logic gate. Since logic gates are building blocks of computers, making a quantum logic gate that could process individual qubits was an important advancement toward a quantum computer. 

The researchers created a quantum version of what is known as a controlled-NOT logic gate. In a controlled-NOT gate, the initial value of the first bit can affect the final value of the second. If the value of the first bit is 1, then it flips the value of the second bit (for example, from 0 to 1). So the possible inputs of the two bits are 00, 01, 10, and 11, and the corresponding outputs are 00, 01, 11, and 10.

A traditional logic gate, with classical bits, can only process one of these inputs at a time. But quantum logic gates work with qubits, which can be in superpositions of 0 and 1, and can therefore process all four possible inputs simultaneously. As a result, a quantum logic gate is capable of performing calculations that are impossible for classical logic gates.

The quantum logic gate prototype introduced in the 1995 paper immediately captured the interest of other researchers, who have cited this work more than 1,200 times to date. Wineland shared the 2012 Nobel Prize in Physics “for ground-breaking experimental methods” that included “the very first steps towards building a new type of super-fast computer, based on quantum physics.”

Scientists started thinking about scaling up this individual quantum logic gate — with its two qubits — to multiple quantum logic gates in quantum computers containing many qubits. With every additional qubit, a quantum computer can manipulate exponentially more complex superpositions. Researchers imagine that quantum computers would be good at doing certain tasks, from factoring the very large numbers used in traditional encryption to carrying out simulations of the quantum world.

NIST Helps Launch an Industry

It’s not a stretch to say that NIST, by making the first quantum logic gate and hosting the first workshop on quantum computing, helped launch the quantum industry.

Many companies, from global corporations to small startups, are now working to create quantum computers. And many NIST alumni, including Monroe, have gone on to work for or even launch some of these companies.

There are multiple ways to make quantum computers, but those that use quantum logic gates have remained a leading design. Today, there are gate-based quantum computers with upward of 1,000 qubits.

And from the first days of quantum information science, researchers at NIST have provided training and mentorship to many postdocs, undergraduates, early career scientists and visiting researchers who have gone on to do big things in the field.

Though quantum computing has made steady progress, there is still a long road ahead until we have working quantum computers that do useful things that are impossible for classical computers to accomplish, such as realizing the original dream of running Shor’s encryption-breaking algorithm.

“We're still in the early stages of a marathon," said Wineland. 

And when it comes to this quantum journey, NIST and its joint institutions are in it for the long run.