This article was published in the November 2015 issue of Maximum PC. For more trusted reviews and feature stories, subscribe here.
Is our universe an immense computer simulation? Is everything we perceive an illusion rendered in super-resolution pixels? Are we mere pixels, too? Lately, I’ve seen these questions debated more often in Internet forums, and often by reputable scientists. It’s not just another crackpot conspiracy theory.
Usually, I’m far too busy to waste attention on trending Internet debates, which can seem as pointless as medieval arguments over the number of angels that could fit on the head of a pin. But the “universe simulation hypothesis” intersects my knowledge of computers. And one thing often missing from the debate is a realistic forecast of future computing power.
Many (though not all) simulation proponents assume, essentially, that Moore’s Law is forever. They take for granted that computers will keep getting more powerful on a steep curve and will eventually be capable of simulating a universe as complex as ours appears to be. Whoa, I say.
The view from Maximum PC’s window is definitely real.
In past columns, I’ve described the fallacies that sprout like weeds around Moore’s Law. To recap, Moore’s Law is not a scientific law; it’s an astute observation of semiconductor progress that was first made in 1965 and modified in 1975.
I’ve charted its course using all three common variations (doubling component density every 12 months, 18 months, or 24 months). Analyzing it by any measure, we’re already falling behind the predicted curve.
Similar curves are common in science. It took only 44 years for airplanes to advance from the Wright brother’s wood-and-canvas contraption to the first supersonic flight. At that rate, airliners should be hypersonic by now, but the curve has flattened since the 1960s.
Yes, future breakthroughs (carbon nanotubes, quantum computing, whatever) could breathe new life into Moore’s Law, or even surpass it. Computers will keep getting more powerful for a long time. However, that doesn’t guarantee they will eventually grow powerful enough to simulate our entire universe.
Consider that to simulate a single atom in all its marvelous complexity requires simulating its subatomic particles (such as protons, electrons, neutrons) and the even-smaller quarks within them. Simulating a particle requires computer memory to store its properties, plus computing elements (such as transistors) to express its behavior and its interactions with other particles.
Even tiny amounts of matter would demand enormous computer resources. To simulate an entire universe on this scale would require a computer that’s larger and more complex than the simulation.
Therefore, I reject a fully detailed simulated universe. However, we could be living in a partial simulation. Just as today’s computer games simulate only one or a few main characters in some detail and fake the rest, it’s possible our universe works the same way. Maybe the simulation creates a local instance of subatomic-level detail only when a scientist peers through an electron microscope or operates a particle accelerator.
Or maybe, only one person is a fully detailed simulation for which everything else is either faked or created on the fly. A fully simulated Paris wouldn’t exist until this person goes there, and it’s faked again when they leave. Maybe that person is me. Maybe it’s you. But if our universe is as vast and complex as it appears, the physical limits on computer power tell me it must be real.
Tom Halfhill was formerly a senior editor for Byte magazine and is now an analyst for Microprocessor Report.
From maximumpc
from http://bit.ly/1Zf4JfG
