My wife recently gained a new appreciation for my work. She was trying to transfer family videos from old videocassettes to DVDs using various gadgets and software running on her laptop. Inevitably, though, the process always yielded a blank screen somewhere in the recording. These interruptions occurred because the laptop kept overheating. Only after she placed it on a stack of books with a fan blowing directly on it could the computer handle the job.

Heat is one of the worst enemies of electronics. Sit on the sofa with your laptop and you quickly feel the heat on your lap. Often though, overheating can be hard to diagnose. You may notice random errors occurring no matter what program you're running. This is especially true if you use your computer to play advanced video games, which can really tax the microprocessor and the graphics card. If your machine frequently experiences fatal errors or "the blue screen of death" on such occasions, chances are it has thermal management problems. And overheating doesn't just degrade a computer's performance; it can also shorten its useful life.

The electronics industry has tended to deal with overheating more at the system level—using cooling fans, for instance, to regulate rising temperatures inside computers. Until recently, heat was not considered a major problem to be addressed in the design of the ICs themselves. But higher circuit densities and faster clock speeds are making chips run so hot that manufacturers can no longer ignore the problem. According to the International Technology Roadmap for Semiconductors (ITRS), which reflects a consensus of chip manufacturers worldwide, managing heat generation within ICs will be a crucial issue in developing the next generation of electronics. Growing concern has in turn sparked new research into chip designs and novel materials that would allow electronics to run much cooler, thereby improving their performance and extending their life span.

Heat is an unavoidable by-product when operating any electronic device. Electronic circuits contain many sources of heat, including the millions and even billions of transistors that are routinely packed into modern ICs as well as the interconnects—the labyrinthine connections linking these components together. In the past, packaging engineers, not chip designers, were the ones who dealt with overheating. They would position components so that any excess heat would move first from the die to a heat sink, and then they'd use a flow of air—from a cooling fan, say—to dissipate heat into the surroundings. Still running too hot? Just use a bigger fan. For years, such coping strategies were sufficient. But now, with the electronics industry aggressively shrinking chip features below 50 nanometers and moving toward three-dimensional integrated circuits, the era of the big-fan solution has passed.

One good way to assess a microprocessor's propensity to heat up is by looking at its thermal design power (TDP), which represents the maximum sustained power that must be dissipated when the chip is performing a typical task. With each new generation of microprocessor, the TDP has grown. In the first Pentiums, for example, the TDP was below 20 watts; in the Pentium 4 it reached 90 W. The transition from single-core microprocessors to multicore microprocessors partially addresses the escalating TDP and worsening thermal-management issues, as the new multicore chips gain performance not by increasing clock frequencies but by adding processors. But other big heat issues remain, including the appearance of hot spots within the chip, where heat fluxes can climb as high as 1000 W per square centimeter. What's worse, according to ITRS projections, within the next five years, up to 80 percent of microprocessor power will be consumed by interconnect wiring in regions of the chips that are particularly difficult to cool.