Supercomputer Reveals the Cosmos in a Different Light
When a researcher looks at the universe as a whole, the natural scientific instinct is to wonder, “What does this remind me of?” Apparently, two of the answers are the brain and the internet.
The brain works on a series of neural pathways that constitute a neural network while the internet acts similarly, with bandwidth signals replacing neurons. The universe and how its objects react to each other create a similar network as well. That is, according to research done at the San Diego Supercomputer Center (SDSC) by the Cooperative Association for Internet Data Analysis (CAIDA).
“The discovered equivalence between the growth of the universe and complex networks strongly suggests that unexpectedly similar laws govern the dynamics of these very different complex systems,” said Dmitri Kriuokov, an author of the paper which used Trestles, a data-intensive supercomputer at SDSC, to determine the universe’s causal network.
Essentially, all of the universe’s phenomena in their interactions with other objects constitute the causal network. An example of this is a supermassive black hole pulling large massive objects (like stars) toward it in some orbit, creating a galaxy.
In measuring the universe, a key part to determining which phenomena displayed causation was determining the limitations. For example, if a source of light were to be shone from Earth at t=0, that light has no chance of affecting anything that is at a distance of ct, where c is the speed of light in a vacuum and t is time.
As such, the researchers at CAIDA were able to determine spheres of influence for each particular phenomenon, reducing the amount of objects it could influence and thus significantly shrink the computational requirement. As a result, computational scientist Robert Sinkovits was able to drastically reduce the time it would take to compute the universe’s causal network from a few years to around 30 hours.
It is important to note that the results—the de Sitter causal sets as the computational physicists call them—are more revealing at the edges of the universe and at times much greater than t=0 since there is less intervening matter with which to interact in that scenario. When taking that into account, the data ended up showing a remarkable resemblance to how networks, like those formed by neurons and computer architects, are optimized.
“In particular, de Sitter causal sets have exactly the same graph structure that maximizes network navigability,” according to the paper. The implications are potentially exciting to physicists, as it could help explain the existence of dark matter, a substance that permeates the universe and accounts for a large majority of its mass, but cannot be pinpointed.
Of course, the larger implication that networks of all types could be fundamentally tied to the same laws is also exciting. While Kriuokov was quick to note that this by no means implies that the universe is a giant brain or vice versa; that this correlation exists at all implies…something (it is not yet known exactly what) about how complex things are formed. From a physics standpoint, at least, it is highly improbable that this correlation exists by accident.
“The most frequent question that people may ask is whether the discovered asymptotic equivalence between complex networks and the universe could be a coincidence,” said Krioukov. “Of course it could be, but the probability of such a coincidence is extremely low. Coincidences in physics are extremely rare, and almost never happen. There is always an explanation, which may be not immediately obvious.”
Full story at UCSD News Center