How to Build a Quantum Computer
UVM physicist wins NSF CAREER grant to study entanglement
- By Joshua E. Brown
Adrian Del Maestro wants to kill fluffy bunnies. Now that I have your attention, be forewarned that the following short story is about quantum physics, the end of Moore’s Law as we know it, and what Einstein’s “spooky action at a distance” might have to do with some futuristic cousin of your iPhone. However, no actual bunnies will die.
Instead, Del Maestro, assistant professor of physics, has won a prestigious 5-year CAREER grant from the National Science Foundation to study entanglement—that bizarre reality of atomic particles where measuring, say, one photon in an entangled pair instantly determines the state of its partner particle, even if they are miles apart—and how entanglement might be applied to create a new generation of ultra-fast quantum computers.
A traditional computer relies on bits. A classical bit is either one or zero. The transistors in a computer circuit board represent this one/zero binary by either being on or off. A very large string of ones and zeros is the foundation of all the codes that make a computer work. "Moore's law" is an observation that, since the 1960s, the number of transistors that can be packed onto a circuit board has doubled approximately every two years. In 1971, 2,300 transistors could be packed onto an Intel computer chip. Last year, that count had risen to over 5.5 billion transistors on a commercially available chip.
But Moore’s law is running into physical limits—“quantum limits,” Del Maestro says. “If you make the distance between the terminal so small, then electrons can tunnel, quantum mechanically, through the barrier. Then you don’t know if it’s one or zero, and you start to get many errors. It could be a zero or a one—at the same time—when things get that small,” he says—and you’ve reached the ultimate size limit of a traditional silicon transistor.
“So that's a problem,” Del Maestro says. “But that can also be looked at as a huge opportunity.”
“Instead of looking at that ‘zero or one?’ question as a problem, maybe we can rethink computation as a way to use that uncertainty—to use entanglement as a resource,” he says. The trick, Del Maestro says, is how to avoid what, even in the scientific journals, is called “fluffy bunny” entanglement.
“Entanglement is the fundamental property of quantum mechanics,” he says. In a rough sense, it’s the fact that when a group of particles are mixed together into a system they maintain connections, even after the parts are physically separated. In an experiment in the Netherlands, reported last fall in Nature, scientists entangled electrons and then sent them in opposite directions for almost a mile. Measuring the spin of one of the electrons instantaneously determined the outcome of the measurement of the spin of its partner.
But the mile isn’t the point. Even if they are on opposite sides of the galaxy, they are entangled. This is what Einstein called “spooky action at a distance,” and though it might seem to violate the laws of the universe it really just shows that our human view of location is an illusion. Light speed is the ultimate speed limit and classical information can’t go any faster than that, but as George Musser has written, in the tiny world of quantum mechanics, “there may be no such thing as place and no such thing as distance.”
And it gets weirder. Until they are measured, atomic particles can be in what physicists call a “superposition,” meaning that the particles can be in all possible states—at once. Indeed this paradoxical truth (that Schrödinger made famous in 1935 with his both dead and alive cat) is a necessary foundation of entanglement. This is what Del Maestro means by the electrons in the transistor being a one and zero—and millions of possibilities in between—at the same time.
And it is this unmeasured probabilistic condition that Del Maestro and other theorists see as the engine for a fundamentally new kind of computer—a quantum computer. Instead of relying on a binary bit, these computers will have qubits—quantum bits—as their base unit. A qubit might be one of those unmeasured electrons. Instead of being just a one or zero, a qubit can be in multiple possible positions at the same time. In essence, a qubit can store and consider multiple possibilities simultaneously—which, in theory, could exponentially increase the speed of a computer. “Basically, you can have much more information in a quantum bit,” Del Maestro says. “This CAREER project is learning how to use that information.”
Now imagine a bunch of, say, atoms of helium cooled to near absolute zero. At this low temperature the atoms form a strange puddle called a “superfluid” where the puddle is really a pile of entangled atoms all sharing a superposition. That’s a lot of entanglement. “This could be the fuel for a quantum computer,” Del Maestro says.
But is all this entanglement—what physicists call “many-body” entanglement—just like a fluffy toy bunny at the carnival—very enticing but ultimately useless? Isn’t it naïve to chase after particles that can’t be distinguished from each other or properties that can’t be measured? Until last spring, a number of prominent physicists would have said: probably yes. But in an amazing experiment announced in April 2015, a team at Harvard was able to make real-world measurements of the amount of entanglement in lattices of these ultra-cold atoms. This, and other related recent discoveries, “brings the technological exploitation of many-body entanglement as a resource within reach,” Del Maestro notes.
Del Maestro’s pioneering work—including his invention of the first theoretical method to measure “operational entanglement” in a many-body quantum system—will complement these experimental efforts. With his new support from the NSF, Del Maestro and his students will spend the next five years exploring the mathematical foundations of entanglement in quantum liquids and ultracold atomic gases. They’ll be developing algorithms on conventional supercomputers—including processors on the Vermont Advanced Computing Core at UVM—that seek answers to questions like: how much entanglement can be extracted from a superfluid—the wildly complex fuel of a quantum computer—and transferred to a more-orderly register of qubits, say a lattice of electrons?
Instead of being the purview of quasi-philosophical speculations, quantum entanglement (that now can be easily created in a modern laboratory) may soon be used in the macro-world of human society—as a tool for information processing, secure communication, and computers many millions of times faster than today’s fastest. It can “drive the next technological age,” says Adrian Del Maestro.