The semiconductor industry has been dominated by "Moore's Law" for decades. Every time it seems we're about to reach the threshold beyond which chip size and density can't possibly go any further, some new breakthrough prolongs the lifetime of the silicon chip just a little bit longer.

Too bad we're not making comparable strides in the energy sector, because without sufficient energy, how will those sturdy little silicon chips be able to run? Kicking off the 2007 Industrial Physics Forum with an overview of the energy landscape, MIT's Mildred Dresselhaus recommended that we "Think big and go small," and called for "a Moore's Law" for energy efficiency. "A few percent in improvement means nothing" in the grand scheme of things," she insisted: "We need an order of magnitude improvement."

The numbers she cited are quite sobering. Global energy demand is expected to double by 2050, and triple by 2100 -- faster than the world population as a whole, which stood at 6.5 billion in 2005 and is expected to reach 8.9 billion by 2050. The energy demand growth rate is so steep, there's no way we'll be able to meet those needs if we continue to draw 85% of our energy from fossil fuel resources. Those resources are projected to peak around 2037, at which point we will have used up some 90% of known oil reserves. It's just not sustainable.

The good news is, there are emerging alternatives, including nuclear power, solar, wind, biomass, and geothermal energy strategies. The bad news is, those alternatives aren't advancing fast enough in terms of meeting the necessary targets of energy output, cost efficiencies, and the like.

Dresselhaus emphasized repeatedly in her talk that no single energy solution will suffice; we need a diversity of sources to close the gap between available sources and our future energy needs. But her comments just on the solar power option alone illustrate the magnitude of the energy challenges we face. "Solar energy is very plentiful, but we're not using much, and [the existing technology] is not competitive on a cost-efficient basis," she said.

The sun deposits mind-boggling amounts of energy to Earth every day: a whopping 1.2 x 10<5> terawatts. The 7.8 magnitude earthquake that leveled San Francisco in 1906? The energy of that event is equivalent to a mere one second of sunlight. A day and a half of sunlight is equivalent to 3 trillion barrels of oil. So why doesn't everything run on solar power? Sadly, harvesting solar power is not that simple, because there are always energy losses to contend with when harvesting the sun's energy for practical purposes. It's basic thermodynamics, the Great Cosmic Killjoy of physics. Today's solar cells operate at about 32% efficiency; with a bit of ingenuity, Dresselhaus thinks we can get them up to 50% efficiency, but even to achieve that much will require truly revolutionary advances in photovoltaics.

Solar cell technology is also more costly, at the moment, than fossil fuels, which might explain why photovoltaics currently only account for .0002 terawatts of world energy needs; the goal is to reach 1.5 terawatts by 2050. To do that, said Dresselhaus, we will need to decrease the cost of photovoltaics by an order of magnitude. We'll also need to figure out more efficient ways of storing the sun's energy until it's needed for use. And that's where the "go small" component of her strategy comes in, because nanoscale materials will play a critical role. For starters, there is a higher surface to volume ratio in many nanomaterials, the better to promote catalytic interactions to yield usable energy. To that end, scientists are already working on quantum-dot based solar cells, as well as nanostructured thermoelectric materials.

Dresselhaus also name-checked such promising sustainable energy sources as hydrogen -- the focus of a major initiative launched in 2003 by the Bush administration -- thermoelectric conversion, modifying the biochemistry of plants and bacteria to produce more efficient biofuels ("designer plants for designer fuels"), artificial photosynthesis, and solid state lighting. The latter, she said, has come the closest to achieving a kind of "Moore's Law' in the energy sector in terms of progress in energy efficiency and cost competitiveness, although some challenges remain.

Policy-wise, she called for road maps setting out the challenges and target goals for each energy source, and emphasized repeatedly the need to promote international collaborations. "Energy is a big, complex system," she concluded, and we need to mind the gap between existing resources and future needs. If we want our ever-smaller computers to keep running, we'll need to look to a more sustainable mix of energy resources. Starting... now!