Skip to main content

Scientists use a supercomputer to build a simulation of the known universe

map of the universe supercomputer euclid flagship
J. Stadel
What do you see when you look at the above picture? If you say, “the inside of the grossest hair-clogged drain ever seen” you would be wrong.

In fact, this hairy-looking Rorschach test is an image taken from a project at the University of Zurich in Switzerland — where researchers have used a supercomputer to create a simulation of our entire known universe — made up of some 25 billion virtual galleries. Stretching a billion years across, the map shows how dark matter is distributed in space, with the yellow clumps showing dark matter and the white areas being the cosmic void, aka the lowest density regions in the universe.

Recommended Videos

The resulting catalog will be used to calibrate experiments on board by the Euclid Consortium, a collaboration with more than 1,000 scientists in the world, whose aim is to map out billions of galaxies in the cosmos and decipher the nature of dark energy. The information will ultimately be used for researching the universe with the Euclid satellite, which will be launched into space in 2020.

“The techniques we have developed are unique: A fast multipole method for gravity that works on thousands of nodes in parallel, and the use of graphics cards to accelerate the calculation,” Romain Teyssier, a professor of computational astrophysics, told Digital Trends. “Because of these two innovations, our code has achieved the largest simulation so far of the entire observable universe, featuring 2 trillion digital particles. We used the Piz Daint supercomputer of the Swiss national supercomputing center, which is made of more than 5,000 GPUs. The unique combination of our innovative code and this world-leading machine is really making this result exciting.”

Executing the code took 80 hours for everything to compile. To put that number in perspective, the Piz Daint supercomputer — a Cray XC30 system — can compute in one day more than a modern laptop could compute in 900 years. Theoretically, that means that trying to generate this project on your run-of-the-mill notebook would take roughly three millennia.

The work is described in a paper published in the journal Computational Astrophysics and Cosmology.

Up next? Things could potentially get even more exciting. “Our goal is to prepare the next simulations, and include more physics like massive neutrinos, or simulate a universe with different laws of gravity,” Teyssier said.

Luke Dormehl
Former Digital Trends Contributor
I'm a UK-based tech writer covering Cool Tech at Digital Trends. I've also written for Fast Company, Wired, the Guardian…
Intel and University of Pennsylvania are using A.I. to find brain tumors
Intel Meltdown

Intel and the University of Pennsylvania's Perelman School of Medicine have teamed up to create artificial intelligence models that can identify brain tumors.

“A.I. shows great promise for the early detection of brain tumors, but it will require more data than any single medical center holds to reach its full potential," said Intel Labs principal engineer Jason Martin in a statement.

Read more
Self-driving vehicle rules set to loosen under Trump, report says
self driving looser rules trump screenshot 2024 10 at 54 56 pm 6708947b14810

Tesla “has been very clear the future is autonomous,” CEO Elon Musk said in October, shortly after unveiling the Cybercab, Tesla’s self-driving robotaxi.

It now seems that Musk, who was recently nominated to lead a newly-created "Department of Government Efficiency," is sharing his crystal ball with the incoming Trump administration.

Read more
Honda doubles down on ‘holy grail’ of EV batteries
honda solid state battery production first electric suv 3

While some automakers are scaling back their production of electric vehicles, Honda is basking in the glow of a successful launch of its Prologue EV in the U.S., and was recently dubbed “North America’s most committed automaker.”

And now, Japan’s third-largest automaker is showing a similar commitment to making EVs more efficient and affordable, zeroing in on the production of its own in-house solid-state batteries, also known as the ‘holy grail’ of EV batteries.

Read more