A billion-lightyear section of the simulated universe. The yellow clumps are dark matter concentrations, and white areas are cosmic void.
A billion-lightyear section of the simulated universe. The yellow clumps are dark matter concentrations, and white areas are cosmic void. Joachim Stadel, UZH

Today, we understand that the universe is expanding, but back in the 1920s, even Albert Einstein thought the universe was static; he even altered his famous theory of general relativity to make it work in a fixed universe. While we now know other galaxies are drifting apart faster and faster, we’re still not entirely sure why. So, in 2020, the European Space Administration will launch Euclid, a satellite that will spend six years mapping the cosmos. But before Euclid looks into the depths of time and space, scientists need to figure out how to process the mountain of data it will collect. To do that, they’ve created a gigantic chunk of a mock universe using a supercomputer in Switzerland.

It took 80 hours for the computer, known as the “Piz Daint,” to generate two trillion dark matter particles, which then formed into 25 billion simulated galaxies that occupy a virtual space the same size as the one Euclid will survey for real. Euclid’s observation of galaxies and how they distort light will help scientists learn more about dark matter and dark energy—the forces believed to be driving universe expansion.

This simulated universe evolves, over around 10 billion virtual years, into the clusters of galaxies on the left.
This simulated universe evolves, over around 10 billion virtual years, into the clusters of galaxies on the left. J. Carretero (PIC), P. Tallada (PIC), S. Serrano (ICE) and the Euclid Consortium Cosmological Simulations SWG

The model galaxy catalog will be distributed to more than 1,000 scientists involved with the Euclid mission so they can experiment with the best ways to collect and process data. They will work out a plan for who will analyze what, because when you’re mapping the universe, practice makes perfect.