IBM Quantum Challenge 2021: Here’s What to Expect
By Ryan F. Mandelbaum
On May, 20th at 10:00 am EDT, we’ll be kicking off the next IBM Quantum Challenge to celebrate the history of quantum computing.
This year is a special one — it marks the 40th anniversary of the Physics of Computation Conference, held at MIT’s Endicott House, where some of the era’s foremost thinkers began to seriously consider a computing device based around the mathematical rules of quantum mechanics. It’s also the five-year anniversary of IBM Quantum putting a quantum computer on the Cloud for anyone to use, which directly led to the formation of the global, open-source Qiskit community centered around programming quantum computers. The challenge will celebrate these anniversaries by featuring five quantum computing exercises, each surrounding a facet of quantum computing’s history.
Quantum computing has come a long way since Richard Feynman uttered that “Nature isn’t classical, dammit, and if you want to make a simulation of nature, you’d better make it quantum mechanical.” In the 40 intervening years, physicists have blueprinted architectures for how such simulators would work, devised algorithms with which a quantum computer could beat a classical computer, built physical hardware, discovered error correction schemes for that hardware, and even started to realize Feynman’s vision. Below, you’ll find some background on each of the IBM Quantum Challenge’s problems with their importance to quantum computing history. At the end of each section, you can also find relevant learning materials to prepare.
1980: Toffoli Gate
Among the themes of the 1981 Physics of Computation Conference was how harnessing the properties of physical systems might improve computer processing. In particular, the idea of overwriting or deleting data in a computation is very different from the way that physical systems evolve over time in the real world. This motivated research into a new approach, known as reversible computing, which offered the possibility of more energy efficient devices. In thinking about this theme, Tommaso Toffoli devised a reversible version of classical computing’s AND and NAND gates — now called the Toffoli, or controlled-controlled-NOT gate. This is a universal gate for classical computation, meaning that any program can be built from many instances of this gate.
The idea of a physically motivated approach to computation soon went beyond a just a reversible form of classical computation. By harnessing the unique behavior of quantum physics, physicists realized that they could build a completely new computational paradigm. Drawing on the existing work on reversible computing, the quantum implementation of the Toffoli gate became one of the first building blocks for quantum computing. Though not universal for quantum computation on its own, its classical universality makes it very important for tasks such as building quantum oracles.
The IBM Quantum Challenge’s first problem introduces the idea of a logic gate — an operation that acts on bits — and extends it to quantum bits. Quantum gates string quantum bits into circuits, the core unit of quantum computation. This problem covers each of the most important gates, and how to combine them in order to create the Toffoli gate.
You can learn the basics of quantum gates in Chapter 1 and 2 of the Qiskit textbook and Lecture 1 to 3 in Qiskit Global Summer School (QGSS) 2020. If you are new to Qiskit, check out the Hello World episode of the Coding with Qiskit Season 1 to stat writing you first quantum program!
1994: Shor’s Algorithm
The conversations at Endicott House may have gotten mental gears turning, but quantum computers still lived mainly in the realm of thought experiments — until Peter Shor debuted his namesake algorithm. The algorithm uses quantum circuits in order to solve a period finding problem, or, how many times a function repeats in a given period of time. But it also allows the solver to factor numbers exponentially faster than any classical computer can. Shor’s algorithm proved that a quantum computer could have important applications beyond simulating physics, and would be a worthwhile direction to pursue for computation overall.
Quantum computers large enough to factor large numbers — such as the large numbers with which the ubiquitous RSA encryption scheme encodes data — are likely many years away. However, Lieven M.K. Vandersypen, Matthias Steffen, Gregory Breyta, Costantino S. Yannoni, Mark H. Sherwood and Isaac L. Chuang were able to demonstrate hardware factoring the number 15 as early as 2001, and you can use a quantum computer to factor small numbers with Shor’s algorithm today. The IBM Quantum Challenge’s second problem requires you use Shor’s algorithm to factor the number 35 on a real device.
For a quick introduction to Shor’s algorithm, you can watch Episode 7 from Coding with Qiskit season 2, or read Chapter 3.7 of the Qiskit textbook. If you want to gain a deeper understanding, Lecture 7 to 12 in QGSS 2020 will walk you through the building blocks of Shor’s algorithm, namely quantum Fourier transform, quantum phase estimation, and the period finding problem.
1995: Quantum Error Correction
Shor’s algorithm gave quantum computers a worthwhile use case — but the inherent noisiness of quantum mechanics meant that building hardware capable of running such an algorithm would be a huge struggle. Classical computers typically include redundancies in their bits in order to avoid errors. Quantum computers are finickier, however — measuring a qubit destroys its quantum state, and the no-cloning theorem prevents you from making an exact copy of a quantum state. But in 1995, Shor released another landmark paper: a scheme that shared quantum information over multiple qubits in order to reduce errors. Today, error correction schemes are one of the most important areas of quantum computing research and development.
Error correction relies on the idea of the error correcting code, or a strategy of encoding multiple physical qubit values into one overall qubit value. The third problem requires not only implementing this code, but thinking about error correction in the way that quantum researchers do, by tailoring the error correction scheme to a specific quantum processor.
You can learn the basics of quantum error correction using a simple repetition code in Chapter 5.1 of the Qiskit textbook and Lecture 13 to 15 in QGSS 2020.
2007: Transmon Qubits
Following Shor’s error correction paper, scientists started looking for systems that showed controllable quantum behaviors to use as the quantum computer’s qubits. Some teams pursued powerful magnets probing molecules’ spin states through Nuclear Magnetic Resonance, the same technology behind MRI machines. Others used ions trapped by lasers, photons traveling through optical setups, or the spin states of electrons. But in 2007, a team of physicists based at Yale debuted a quantum processor based on the behavior of electrons traveling through circuits of superconducting wire, impeded by an element called a Josephson junction.
The Josephson junction-capacitor circuit acts like an oscillator, and holding the superconducting circuit in cold temperatures quantizes the current — it only takes on discrete modes or linear combinations of discrete modes. Pulses of microwave energy initialize the qubits into the lowest state, moves them into the excited state, create superpositions of the two states, or serve as gates that entangle the qubits with others. The Josephson junction introduces a nonlinearity element, creating an uneven spacing to the energy required to move between each mode and allowing the programmer to use only the bottom two modes.
Transmon qubits serve as the backbone of IBM’s quantum systems today. The fourth problem offers a deeper understanding of how they work, allowing you to perform experiments on real quantum hardware using Qiskit Pulse.
You can learn more about superconducting qubits in Chapter 6 of the Qiskit textbook and Lecture 16 to 21 in QGSS 2020. For learning Qiskit Pulse, you can watch this tutorial video by Qiskit Pulse developer Lauren Capelluto.
2014: Variational Quantum Eigensolver
During the last decade, quantum computers matured quickly — and began to realize Feynman’s initial dream of a computing system that could simulate the laws of nature in a quantum way. A 2014 paper first-authored by Alberto Peruzzo introduced the Variational Quantum Eigensolver (VQE), an algorithm designed to find the ground state energy of a molecule with much shallower circuits than other approaches. And, in 2017, the IBM Quantum team used the VQE algorithm to simulate the ground state energy of the lithium hydride molecule.
VQE’s magic comes from outsourcing some of the problem’s processing workload to a classical computer. The algorithm starts with a parameterized quantum circuit called an ansatz (kind of like a best guess), then finds the optimal parameters for the circuit using a classical optimizer. The problem’s advantage comes from the fact that a quantum processing unit can represent and store the problem’s exact wavefunction, an exponentially hard problem for a classical computer. The fifth problem of the IBM Quantum Challenge allows you to realize Feynman’s dream yourself, setting up a variational quantum eigensolver to simulate a large molecule.
You can learn more about VQE for simulating molecules in Chapter 4.1.2 of the Qiskit textbook and Lecture 22 to 27 in QGSS 2020. We have recently released the Qiskit Nature application module to allow researchers in different areas of natural sciences (including physics, chemistry, material science and biology) to model and solve domain-specific problems using quantum simulations. You can find the tutorials here.
We hope you enjoy this trip through quantum computing’s history with this year’s IBM Quantum Challenge. Good luck, and get started by clicking here.
