The History of Games for Quantum Computers

It has been over a year since the first ever quantum computer game was created. It is time to write their history so far.

March 2017

It began in the bygone age of 2017, a man had an idea: To make a game for a quantum computer.

Hardware wouldn’t be an issue: IBM had provided that through the IBM Q Experience. The software was also there: Project Q, a quantum SDK capable of running jobs on IBM’s devices, had recently been released. All that was needed was a game.

The game chosen was Rock-Paper-Scissors. It was adapted to suit the strengths of qubits, and the first quantum game was made. It was called Cat-Box-Scissors.

It wasn’t very good, truth be told. Just a simple first experiment. Little more than a random number generator with a story. Quantum computers deserved something better.

And so, only a little later, came the first multiplayer game made with a quantum computer. With techniques once used to probe fundamental properties of the universe, this game would play Battleships.

April 2017

These first quantum computer games were simple examples, limited to the command line. They ran on the quantum processor in real time, which meant waiting in the job queue. Unfortunately, waiting for snippets of text is few people’s idea of fun.

So what about a game based on preexisting data? The quantum computer can generate everything needed beforehand, allowing the game to run in the fast and responsive way we all expect from modern programs.

From this idea came Quantum Solitaire, which was exhibited at Fantasy Basel (the Swiss answer to Comic Con).

May 2017

May brought two quantum games. One made with the same philosophy of Quantum Solitaire, running on preexisting data. It was inspired by Hunt the Wumpus, an early dungeon crawler. This was Hunt the Quantpus.

It also brought a new version of Battleships. This was another game to run in real time. But rather than using Project Q to handle the software, as before, it used QISKit: IBM’s newly released native SDK for their hardware.

This was the first real-time quantum game to run for multiple rounds. As such, it was the first to go beyond just being a fancy random number generator. It was also created with a noble purpose: To help people to learn quantum programming.

Due to its increased sophistication, all previous games were downgraded to being mere experiments. Battleships with partial NOT gates was the true first quantum computer game.

May 2017 is also notable for being the month that IBM announced their 16 qubit processor. Like all cloud based quantum processors, it was destined to one day play games.

June 2017

Every game so far was designed by just one guy. Me!

In June of 2017, that changed. Rigetti, a quantum computing startup, released their own quantum SDK. At the same time they also made a simple game, aimed at providing a little demonstration of quantum computing. Check it out on their website here.

It doesn’t strictly count as a quantum game, since it uses a simulated quantum computer rather than a real one. But Rigetti have the hardware and software to implement it on a real device, so it certainly deserves a mention.

A similar simple demonstration was also made by IBM back in March 2017 (see here). Though this one arguably doesn’t quite count as being a game, and never claims to be one, it deserves a mention too.

I was still developing my own quantum games, of course. June brought a new entry into the long-running Quantum Battleships franchise. As a further tutorial in quantum programming came Battleships with complementary measurements.

June also had the BlueYard Quantum Leap event: a meeting of researchers, startups, investors and journalists all interested in quantum computing. I managed to wangle a ticket as the guy who makes quantum games. There I had a chat with Will Zeng of Rigetti about Spacewar!, one of the first games for normal computers. I also talked with Alan Ho from Google about some thoughts I had on their proposal for ‘quantum supremacy’, and with Jerry Chow of IBM about their 16 qubit device. These conversations went on to inspire some of what was to happen in August.

July 2017

In July, a section was added to the tutorial for the QISKit quantum SDK. It had been promised for a while, but in this month it was finally founded with a notebook on the ‘Quantum Counterfeit Coin Problem’.

This isn’t really a game in the standard sense. It’s something that you can study with game theory, but not really something that you play. This is also true of another early addition to this section: Quantum Magic Squares.

Nevertheless, this section of the tutorial was intended to contain games that can be played as well. Accordingly, it soon became the home of Battleships with partial NOT gates.

August 2017

Back in July of 2016, a bunch of mostly Googlers published a paper. It was a plan on how to prove what they call ‘Quantum Supremacy’: That quantum computers can be better than normal ones at some tasks.

The eventual demonstration of this likely won’t be for a useful task. It will be for something quite abstract, that is heavily biased in favour of the quantum contender. It will also take a good decade or so after the first claims of ‘supremacy’ for true quantum computers to emerge.

Nevertheless, the more interesting we can make the task, the more interesting the result will be. The more relatable we make the task, the more understandable the result will be. So let’s make it into a game!

That was the idea I first had back at the start of 2017. It slowly developed over the first half of the year, and almost got abandoned. But the conversations I had at the BlueYard event in June focussed my thinking.

In August, it emerged. Quantum Supremacy in game form: Quantum Awesomeness.

Like Spacewar! before it, this was a game design to use the hardware to its limits. It would be able to provide context for a supremacy result. It would also provide benchmarking data for devices too small or noisy for supremacy.

In this game, the size and connectivity of a quantum processor is presented in the relatable form of a puzzle. Imperfections and noise became an increase in difficulty. With this game, players from any background could start to understand what current devices are really like.

The first version was hardwired for a particular device: IBM’s 16 qubit processor. So that’s where we got the first results.

August was also a big month for Battleships with partial NOT gates. It became part of the QISKit tutorial section for games, as created in July.

A version adapted to be a bit more playable was also created, with players given some text to read while they wait in the queue. A playthrough of this was recorded and put on YouTube. Though intended only as something to be shown to a few people (and so made without narration or background music), it has been viewed by a few thousand brave souls.

The reason for this more playable version was to be part of an event in Aarhus. This was hosted by ScienceAtHome, who make great games about quantum computers.

August then ended with the first conference talk regarding games running on quantum computers. This happened at Gamescom in Cologne, one of the world’s biggest trade fairs for gaming.

September/October 2017

The Autumn months of 2017 were a quiet time for my efforts towards quantum games. I’d like to say that it was because I was busy with serious science stuff. But that would only be partially true. I also made a superposition of emoticons as a quantum version of Hello World.

It wasn’t so quiet for others, though. A group of people at the University of Osnabrück made a game for a Comparative Machine Learning class. This didn’t just have quantum computing, but neuromorphic computing too!

November/December 2017

Quantum Awesomeness was designed so that any device could play it. No matter what architecture, or size, or connectivity. As long as you had a bunch of qubits, you could play Quantum Awesomeness on them.

Though this was the theory, in practice the software only supported IBM’s 16 qubit device. It was time to widen the net.

At the end of November, a bit overhaul was committed to GitHub. It could now play on the newly upgraded 16 qubit IBM device. It could play on their 5 qubit devices too. In each case, it used IBM’s QISKit SDK.

Another big change was to include support for Project Q, which had been neglected by quantum game development for a few months. Support for Forest by Rigetti was soon added, inspired by their announcement of a 19 qubit processor. Quantum Awesomeness was among the first in the queue to run on this new device, with the first data coming in just before Christmas

All data collected so far (from 3 IBM devices and and 1 Rigetti device) was then put up on GitHub. This allowed players to play Quantum Awesomeness games from all these quantum processors, without the need for direct access. If anyone wondered whether Rigetti’s 19 qubits were better than IBM’s 16, they could find out for themselves. Just by playing a game.

January/February 2018

Quantum Awesomeness continues in 2018. The project improves, and is used as the basis of articles to explain the current state of quantum computing. This includes text-based ‘Let’s Plays’ of games run on different devices.

In late February, it finally became possible to play the game in a browser. No more cloning repos or configuring Jupyter required.

These months also featured the first new game in a while, though it is technically more of a gamified tutorial. Through figuring out puzzles, a player can get their first taste of quantum computing.

The tutorial itself runs on a normal computer. But it doesn’t end there! The program has additional modes that allow programs to be written and then run on a real device. So it just about counts as a game that runs on a quantum computer, even though neither the ‘game’ nor ‘runs on a quantum computer’ claims are very strong.

March 2018

The end of March was the deadline for IBM’s Teach Me QISKit award. It challenged entrants to create interesting Jupyter notebooks, to help teach others the basics of quantum programming. The winner, a simulation of the Ising model, was an excellent example of using quantum computers for scientific purposes. But another entrant was a simple game based on the exotic properties of quantum correlations.

April 2018

On the 6th and 7th April, Rigetti held the world’s first quantum hackathon. Or at least the first to allow programs to be run on real quantum hardware. There were participants from a wide range of backgrounds, doing all sorts of projects. Two teams even made games. One was done by myself and my team mates, Jonathan DuBois and M. Sohaib Alam. It was called Link to the Quantum.

Another was a fun mobile game, made by a team with Rigetti’s own Will Zeng.

May 2018

Since summer 2017, I had been working on a game in collaboration with IBM Research. The idea to let people create jobs to run on real devices, and do it via a mobile game. In the end we created a puzzle game to give people their first task of quantum programming: Hello Quantum. One version is targeted at casual gamers, and released on mobile devices.

Though it doesn’t run on quantum computers at all, it aims to set players up to create their own programs on the IBM Q Experience. And provides everything you need to reproduce the puzzles and solutions on a real device.

We also released an improved version of the command line variant of the game, which I mentioned above in Jan/Feb.

June 2018

Quantum Awesomeness had long been a game that people could play. But it hadn’t yet been presented as a piece of science. It was time to write a paper about it.

This was written as a study of how the quantum programs run by the game can help us understand and compare prototype quantum devices. Little reference to its nature as a game was made.

One of the selling points of the paper is that it covered all quantum processors available to the public. Then, just as I was putting the finishing touches, Rigetti went and released a new device. An 8 qubit one this time.

Fortunately, they let me have a go on it pretty quickly. So Quantum Awesomeness added a new device to its roster.

The Future

Hopefully more things by you and less by me!


Appendix: Quantum game prehistory

Universe Splitter debuted in 2009. It’s a coin flip app, whose uniqueness comes from the fact that it uses a quantum source of random numbers. This allows it to connect to some of the narrative around quantum physics. Specifically, it presents itself as a way of splitting the universe whenever you want to make a decision.

It’s not a game, but it has gamified elements. And it doesn’t run on a quantum computer, but it does use some quantum hardware. So it is a definite forerunner of games running on quantum computers, and an inspiration to show us what can be done even with just a drop of quantum

In 2016 there were a whole bunch of projects combining quantum and games or gamification. For an incomplete list, see this session of talks about them from early 2017, as well as the Quantum Game with Photons.

Though 2016 brought much quantum gamification, only three projects involved real devices. One was the IBM Q Experience.

Don’t be frightened off by the pumpkin. It’s just a relic from the Halloween reskin. The Q Experience is actually a non-scary way of creating simple programs and running them on real devices. It has graphics and a drag-and-drop mechanic that you might find in a game.

Another was the Alice experiment.

Since it was a project by ScienceAtHome, it definitely included gamified elements along with the science. But it was never considered to be a game. And though the hardware was certainly related to quantum computation, it wasn’t a quantum computer.

The third was the Big Bell Test.

This used human players to generate randomness, which was then used in a set of real quantum experiments.

Something else notable had also come out of the University of Bristol a few years before.

Again, it allowed access to real devices via the cloud. And again it had a relatively game like interface. Along with the Alice experiment and the Q Experience, this was an early example of allowing general access to cutting edge quantum devices. Hopefully, they are the forerunners of many more.

Like what you read? Give Dr James Wootton a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.