For fifty years the smartest people and the biggest power brokers in the world have designed and funded and constructed huge, expensive tokamaks, laser arrays, and other awe-inspiring mega-technology, trying in vain to reach breakeven energy output in a thermal plasma fusion reaction - relying on the random collisions of nuclei in a chaotic, high-temperature confined plasma to produce a small percentage of fusion events.
But in 2008 Yue Shi, a DARPA-funded grad student at Cornell, designed and constructed prototypes of tiny linear particle accelerators on microchips, capable of significantly boosting the energy of input particles. Inevitably, she and her advisor Amit Lal examined the tantalizing possibility that chains or arrays of such semiconductor accelerators could push ionized deuterium nuclei to fusion energies. The math is encouraging--read their 2009 patent application here.
This development has the hallmarks of a breakthrough "small is better" approach that could well bypass the entire "big fusion" debacle. By contrast with the brute force, hard-way approaches of the last five decades, which attempted to heat volumes of fuel plasma to over 150 million degrees, precision manipulation of smaller numbers of fuel particles in a tightly-confining microstructure may reach much higher efficiency levels in a "microchip fusor"--perhaps high enough to produce small net power in a small package, e.g. at a flashlight-battery level. Then we would have a safe, sane, and scalable fusion power cell, capable of being "stacked" by the thousands into commercial power arrays--or vehicle power plants ("Mr. Fusion"?).
It would be ironic (though, somehow appropriate) if the macho big-dumb-fusion effort, long the stronghold of elite male physicists, were trumped by a lady grad student with a clever little microchip. :)