The claws are out over claims that an IBM team has developed a computer simulation of a cat's brain.
At the SC09 supercomputing conference in Portland in November, scientists and engineers from Big Blue's Almaden Research Centre announced a cortical simulator replicating one billion neurons connected by 10 trillion synapses running on a Blue Gene/P supercomputer. The research is funded by the U.S. Defence Advanced Research Projects Agency, better known as DARPA (motto: We created the Internet so you can share cat videos) to the tune of about $21 million.
(DARPA's end game, BTW, is to create a prototype chip that simulates 10 billion neurons connected by one trillion synapses, drawing less than 1 KW, taking up less than two litres of space, and able to lounge around all day licking its stomach. Only one of these specs is not true.)
While Almaden's Dharmendra Modha basked in the IBM-directed publicity spotlight, rival Henry Markram, lead of Ecole Polytechnique Federale de Lausanne's Blue Brain Project, stewed quietly. Actually, he stewed noisily, firing off a scathing letter to IBM CTO Bernard Meyerson, dismissing the cortical simulator as a scam.
The argument, in short: “These are point neurons (missing 99.999% of the brain; no branches; no detailed ion channels; the simplest possible equation you can imagine to simulate a neuron, totally trivial synapses; and using the STDP learning rule I discovered in this way is also is a joke).” Those of you who aren't already having synaptic problems thus far can read the rest of Markram's screed here.
From Sharky's perspective, it seems there are enough moggies lying about and gazing at us disdainfully through cold, hard eyes that we haven't much need to simulate the feline braiin process.I can haz cheezburger now?
