Comparing the human brain to the fastest and most powerful computers in the world is a good way to fathom just how huge and complex it is. And the latest research shows, yet again, that even the most badass supercomputers can't hold a candle to the fleshy masses inside our skulls.
On the other hand, they're coming closer than ever before. This month computer scientists from Japan and Germany were able to simulate one percent of human brain activity for a single second using 82,000 processors from the fourth most powerful supercomputer in the world, Japan's K computer.
Here are the numbers: The computer scientists recreated 1.73 billion virtual nerve cells and 10.4 trillion synapses, each of which contained 24 bytes of memory. The simulation took 40 minutes of real, "biological" time to produce one virtual second.
Billions and trillions of simulated neurons and synapses is nothing to sneeze at, but keep in mind how that equates to only one percent of what's going on in our noggins. The brain, by comparison, consists of about 200 86 billion neurons linked together by trillions of synapses, making for a total of hundreds of trillions of different pathways that brain signals travel through. That's a lot of electrical impulses shooting through the brain at once, which means a hellova lot of machine power.
The K computer held the crown for world's fastest computer in 2011, but was recently dethroned by China's Tianhe-2, also known as the MilkyWay-2. The brain simulation was run by the popular open-source NEST software at the RIKEN program for Computational Life Sciences.
For artificial intelligence enthusiasts, the experiment should be encouraging. "If petascale computers like the K computer are capable of representing 1 percent of the network of a human brain today, then we know that simulating the whole brain at the level of the individual nerve cell and its synapses will be possible with exascale computers, hopefully available within the next decade," lead researcher Markus Diesmann said in a news release.
For some perspective, a petascale computer has the combined memory power of about 250,000 regular PCs; an exascale machine would be a thousand times faster than that.
Attempting to replicate the entire brain is so ambitious it's controversial, though that's not stopping Henry Markram, the man behind the $1.3 billion Human Brain Project funded by the European Union. Markram is using the IBM Blue Gene supercomputer, one of the world's fastest, though he says a full-scale simulation of the brain will require a computer 100,000 times faster. It also calls for a comprehensive, if theoretical, "connectome" map of the 100 trillion neural connections—and that's something we don't have, though some audacious neuroscientists are working on creating one.
What's more, simulating something as complex as the human brain will take more than just sheer processing power. The brain's architecture is nuanced and mysterious, and computer scientists have a long way to go to model its function with a machine.
So, the singularity may not be around the corner, but there are still reasons to be interested in creating a model of the brain. Down the road, it could help neuroscience researchers better understand mental diseases. As of now, the supercomputer simulations aren't providing any new understanding into how the brain thinks, but it's a sign of what's possible in the future.
"It's a bit like building a super-connected motorway network, populated with simulated cars, but not yet looking at how that road network reacts to the holiday road rush," wrote brain researcher Peter Mcowan on Medical Xpress. "But there's no doubt that such giant scale simulations will soon yield answers to mysteries about how our brains operate, how we learn, how we perceive, and perhaps even how we feel."