Will Computers Continue to Improve Exponentially?

As Gordon Moore famously predicted in 1965, the amount of computing power we can inexpensively cram in a computer has grown exponentially for quite some time now. However, after 50 years of steady doubling, there are now doubts about how long this development can still continue. We are approaching the physical limits of silicon-based integrated circuits, and some slowdown is already apparent. Making a few billion tiny transistors without a single defect isn't an easy task, and it's a small wonder we're capable of that. The complexity and scale involved are mind-boggling.

Some futurists like Ray Kurzweil argue that the law of ever-cheaper computing holds across past technologies right down to mechanical computers, and will likely continue to hold on future technologies - the attached chart is modeled after one by him. This might be so on the long term, but the road ahead won't be smooth, and we have already run into a few bumps along the way. Making circuits with 45 nanometer half-pitch a few years ago meant inventing new materials so that electrons leaking due to quantum effects could be prevented. As more nanometers are shaved off, continuing this kind of basic research will be required.

As if making chips work wasn't hard enough, there is also another problem: actually manufacturing them in large quantities. Using laser-printing to make features that are smaller than the light wavelength of 193 nanometers involves some impressive optical trickery. Lasers with a smaller wavelength are still in their infancy, and if they ever become practically feasible, they will use considerable amounts of energy. We might be able to make smaller chips with them, but at a high price premium. The insurmountable laws of physics are attacking the industry from every angle.

But what does the fast development of better computing equipment ultimately mean for the society at large? Many would be happier if they didn't have to upgrade their laptop computers every other year anymore. But all other things notwithstanding, few argue against the advancements computers have brought to science and medicine. Developing new life-saving treatments involves heavy computer simulations almost without exception. Computers are also a major factor in recent cultural developments: mix cheap digital hardware with some raw creativity and the Internet, and out come independent artists in every craft imaginable.

Cheap and efficient computers are the most important enabling technology on every area of human activity. Now we've begun to take steps to make devices increasingly more mobile, more aware of their surroundings, more eco-friendly and also available for people in developing countries. Today's smartphone is a yesterday's computer, professional video camera and a hefty amount of media storage rolled into a pocket-sized package. People are increasingly able to take their creative interests wherever they go, and keep in contact with friends and colleagues. On top of that, we're finding lots of new uses for our talented pocket assistants, like translation of conversations. To take the next steps on this path, the underlying chip industry's progress must not stagnate.