Simulate Something

If you want to see how a model’s behavior changes over time, you can run a simulation. Simulations can be done using models in real life, but often computers are used to run simulations in the virtual world.

Computer simulations can be very simple, running on a single computer in a matter of seconds. Or they can be very complex, requiring a room full of networked computers working over several days.

This image shows the visible matter inside a cube representing 248 million light-years. This was the world's most complex scientific simulation of the evolution of the universe ever performed. Michael Norman, a cosmologist from the University of California, San Diego, ran the simulation for more than 130 hours on 512 processors of the Blue Horizon supercomputer.

This simulation required tracking more than a billion particles and performing calculations in more than a billion cells for more than 3 billion years of simulated time.

Image Credit: Michael Norman, Pascal Paschos, UCSD; Robert Harkness, SDSC

Like individual pages in a flip book, computer simulations preserve snapshots of calculations over time. Each snapshot shows the output of a calculation using a slightly different input value. By putting the snapshots together, the simulation presents an animated view of change over time.

Mathematical models provide the equations and rules that define a process. Because mathematical models use numerical inputs and outputs, they allow you to see very specifically how a system evolves over time.

When you gather the numerical values as the simulation progresses, you are left with a comprehensive set of data showing values at every step of the process.