Molecular computing is a generic term for any computational scheme which uses individual atoms or molecules as a means of solving computational problems. Molecular computing is most frequently associated with DNA computing, because that has made the most progress, but it can also refer to quantum computing or molecular logic gates. All forms of molecular computing are currently in their infancy, but in the long run are likely to replace traditional silicon computers, which suffer barriers to higher levels of performance.
A single kilogram of carbon contains 5 x 1025 atoms. Imagine if we could use only 100 atoms to store a single bit or perform a computational operation. Using massive parallelism, a molecular computing weighing just a kilogram could process more than 1027 operations per second, more than a billion times faster than today’s best supercomputer, which operates at about 1017 operations per second. With so much greater computational power, we could achieve feats of calculation and simulation unimaginable to us today.
Different proposals for molecular computers vary in the principles of their operation. In DNA computing, DNA serves as the software whereas enzymes serve as the hardware. Custom-synthesized DNA strands are combined with enzymes in a test tube, and depending on the length of the resulting output strand, a solution can be derived. DNA computation is extremely powerful in its potential, but suffers from major drawbacks. DNA computation is non-universal, meaning that there are problems it cannot, even in principle, solve. It can only return yes-or-no answers to computational problems. In 2002, researchers in Israel created a DNA computer which could perform 330 trillion operations per second, more than 100,000 times faster than the speed of the fastest PC at the time.
Another proposal for molecular computing is quantum computing. Quantum computing takes advantage of quantum effects to perform computation, and the details are complicated. Quantum computing depends upon supercooled atoms locked in entangled states with one another. A major challenge is that as the number of computational elements (qubits) increases, it becomes progressively more difficult to insulate the quantum computer from matter on the outside, causing it to decohere, eliminating quantum effects and restoring the computer to a classical state. This ruins the calculation. Quantum computing may yet be developed into practical applications, but many physicists and computer scientists remain skeptical.
An even more advanced molecular computer would involve nanoscale logic gates or nanoelectronic components conducting processing in a more conventional, universal, and controlled manner. Unfortunately, we currently lack the manufacturing capability necessary to fabricate such a computer. Nanoscale robotics capable of placing each atom in the desired configuration would be necessary to realize this type of molecular computer. Preliminary efforts to develop this type of robotics are underway, but a major breakthrough could take decades.
By Michael Anissimov