Alyosha Molnar works where silicon and neurons meet.
“It’s likely that what we know about electronic circuits applies to neural circuits. The trick is to find where it makes sense,” he says. “I tell people that two-thirds of my research is on chips and two-thirds is on neuroscience.” Molnar didn’t enter academia directly. After graduating Swarthmore, where he helped design the computer control system for the college’s hybrid-electric entry in the Tour de Sol, Molnar wasn’t sure what career path to take. He took his uncle up on an offer to work as a deckhand on his fishing boat. The fishing was bad that year—Molnar had plenty of time to get through Cornell Prof. Steve Strogatz’ text on nonlinear dynamics—and he decided to put off grad school for a “pre-doc” at Rockwell Semiconductor.
The company quickly spun off Molnar’s division, which became Conexant Systems, and after his boss quit, he was left to his own devices. “I came up with this crackpot architecture idea for cell phone radios,” he says. “All the big wigs said it was impossible, that it would never meet spec, but we did it and it worked.”
Molnar’s future at Conexant, which has sold more than 20 million of the radios, seemed assured, but for him success meant drudgery. “I was going to spend the next few years doing minor modifications and be totally bored,” he says. “The other equally important reasons are that I thought I wanted to teach in the long run, and I had encountered interesting stuff at the intersection of biology and electrical engineering.”
So Molnar left for U.C. Berkeley, where he tried to design a neural interface chip. “It was a total failure because I couldn’t communicate with the biologists. It was pretty disheartening,” he says. “I concluded that the only way to build a bridge with biologists was to become a biologist, so I could be the bridge.”
After taking biology classes, Molnar ended up working in a retinal neuroscience lab where he did his thesis on rabbit eyes. “I started seeing neural circuits that looked just like electrical circuits I’d worked on before,” he says. “We demonstrated that many pathways in the retina maintain a linear response, the same way radio receivers maintain their response in the face of interference.”
Electrical engineers can learn a lot from biology, says Molnar. “The average neuron is 10 million times slower than a modern transistor but neural circuits do just as good a job, so obviously we’re missing something and we don’t have a clue what it is,” he says. “There’s something about the architecture of cortical circuitry that’s very smart in some sense.”
And with its focus on biological engineering, Cornell is just the place for Molnar to make those discoveries. “Everyone I talked to said this is one of the best places for interdisciplinary work,” says Molnar. “And the location is a big plus. My wife and I always dreamed of living in a rural place but thought because I work in a technical field it would never happen, but there’s an exception.”
Prof. Molnar's Web site
