A quant’s jump

Okay, so it might be a stretch to place Robert Gelfond in the scientific pantheon with Werner Heisenberg.

Okay, so it might be a stretch to place Robert Gelfond in the scientific pantheon with Werner Heisenberg.

By Jeffrey Kutler
March 2003
Institutional Investor Magazine

But Gelfond, CEO of MagiQ Technologies, is closing in on a commercial breakthrough in quantum information processing, a branch of computing that draws on Heisenberg’s famously paradoxical quantum physics.

For Gelfond, it’s a leap from a career as a quant trader at D.E. Shaw & Co. and Millennium Partners. As a private investor in the 1990s, Gelfond, who earned bachelor’s degrees in mathematics and economics from the University of Virginia, developed a fascination with quantum computing -- then an untested technique for vastly increasing a machine’s data-processing power. The operative theory is Heisenberg’s uncertainty principle, which says that it’s impossible to locate the precise position of an electron within an atom. By extension, many subatomic phenomena cannot be explained by ordinary physical laws. According to quantum’s magic, a bit of data -- an electronic impulse that registers a 1 or a 0 in a conventional computer -- becomes a quantum bit, or qubit, capable of being both a 1 and a 0 simultaneously. A processor’s power rises exponentially with each qubit.

Unable to find a pure play for a seed investment, Gelfond started New Yorkbased MagiQ in 1999 and raised several million dollars from angels like Amazon.com founder (and former D.E. Shaw colleague) Jeff Bezos. In November Gelfond said MagiQ’s first product, a cryptography device called Navajo, was entering test phases and would be on the market by late 2003. “Most people who have heard about quantum think it’s more futuristic than it is,” says the 43-year-old Gelfond.

Navajo is relatively elementary, with a single-qubit processor. More-revolutionary multiqubit computing, for sophisticated financial modeling and other “rocket science,” may still be a few years away. “We’re working on devices of that kind,” Gelfond says. “But we also have in mind other single-qubit products for which we don’t need further scientific breakthroughs -- just more resources.”

Related