Quantum simulator brings hundreds of qubits to bear on physics problems
By Brian Dodson
May 9, 2012
Physicists at the National Institute of Standards and Technology (NIST) have built a quantum simulator that contains hundreds of qubits - quite a jump from the the 2-8 qubits found in state-of-the-art digital quantum computers. The simulator has passed a series of important benchmarking tests and scientists are poised to study problems in material science that are impossible to model using classical computers.
Many important problems in physics and materials are poorly understood at present - not because their basic physical interactions are unknown, but rather because the quantum mechanical description is highly complex and difficult to solve. For example, even an apparently simple problem such as finding the ground state of the hydrogen molecule cannot be exactly solved. Even when using approximations which are physically fairly reasonable, modern computers cannot simulate quantum systems having more than a handful of interacting particles.
A great deal of scientific and popular interest has surfaced during the last 20 years concerning the use of quantum computers to get around this problem of complexity. Unfortunately, a rough rule of thumb is that you need the same number of qubits as your problem has quantum degrees of freedom to overcome the exponential complexity of quantum systems, and current quantum computers don't yet make the grade.
On to quantum simulators. What is meant by that term, how does one work, and why is the use of a quantum simulator a reasonable alternative to the use of a large-scale quantum computer?
A quantum simulator consists of a collection of quantum mechanical objects, typically atoms with spin, arranged in a simple geometry and with known interactions between the objects. An accurate description of the energy of any particular configuration of the spins must be known, even though the quantum state of the simulator is too complex to write down.
In use, the initial state of the simulator (e.g. which spins are up and which are down) is initially set, and you observe the evolution of the system. Sometimes this will be done as some measure of temperature is varied, and other times as the strength and/or direction of an external field, such as a magnetic field, is changed. Usually the observation is of some average property of the simulator, such as the total magnetic field of the spins in the simulator.
Effective use of a quantum simulator involves describing the physics of a difficult problem in terms of the physics controlling the behavior of the simulator.
On to NIST's new quantum simulator, which consists of a collection of hundreds of beryllium ions in an ion trap. The ions self-assemble into a flat crystal with a triangular pattern less than a millimeter in diameter. The beryllium ions have a spin of 1/2, and the simulator is placed in a strong (~4.5 Tesla) magnetic field perpendicular to the plane of the crystal.
The photo above shows the NIST quantum simulator with the ions located in a single triangular planar crystal. The ions are all fluorescing, telling us that all the ion spins are in the same state. The individual spins are controllable, so that the initial state of the lattice can be set. In addition, the lattice can be placed into a superposition of differing initial states, and pairs of ions can be entangled as well, adding additional depth to the control one has over the behavior of the simulator.
In the simulation, laser beams are used to cool the ion crystal to near absolute zero. Carefully timed microwave and laser pulses then are used to cause the qubits (individual spins) to interact. Controlling the interaction of the qubits can allow an analogy to be made between the simple simulator structure and complex physics defining a much more difficult problem. Although the appearance of the two systems may be very different, their behavior is engineered to be mathematically identical, so that a solution to one yields a solution to the other. Problems that can be studied in this way include gases, liquids, solids, alloys, neural networks, flocking behavior, the interaction of heart cells, and even social behavior.
Given the remarkable progress reported by the NIST researchers, it won't be long before the quantum simulator will be a part of the toolbox of every complex materials physics researcher.
NIST postdoctorial fellow Joe Britton describes the simulator in the video below.