In the race to achieve higher speed computation, conventional lithography-based Very Large-Scale Integration (VLSI) is bumping up against a brick wall that threatens to disrupt Moore’s Law which states that the number of transistors in a dense IC doubles about every two years. That formulation is more of an observation of past performance than an immutable law with permanent relevance. This nonetheless valuable insight has until now been an accurate description of the rate of progress in one area of computing hardware capability. Of course, eventually there has to be an endgame. It seems as though the way things have been going, the endgame is imminent.
An incomplete list of reasons for the slowdown, invalidating Moore’s Law as universally valid, includes limits in ultrathin oxide insulators, short channel problems, difficulties in small-area doping, and the rising cost of undersized lithography. These and other reasons have forced a new approach to IC making.
It surfaced incrementally toward the end of the twentieth century as researchers saw merit in the field of molecular-scale electronics. Two promising avenues in particular seemed worth pursuing as a means for overcoming obstacles preventing the further shrinking of circuitry: the resonant tunneling-diode and transistor, and quantum dot and single-electron devices.
The electronic devices researchers have in mind for the next step in Moore’s Law have dimensions on the order of a single molecule, hence the moniker molecular electronics. To succeed, a new molecular-scale technology would have to include provisions for solving interconnect problems, include self-alignment during fabrication, and overcome the need for super-cooling. The whole thing could be doable with the use of chemically synthesized interconnects, accomplished at minimal expense with routine pre-doctoral lab techniques.
Molecular electronics, if it is to become a reality outside the laboratory, will have to embrace the demands of quantum (sub-100 nm) sizing. This will at once afford amazing challenges and great opportunities. The foremost problem is discovering a way to acquire and maintain contact between molecular components and the much larger-scale electronic circuitry outside that regime. The foremost opportunity is the potential for quantum acceleration of the computation process. At such close quarters, quantum phenomena predominate.
Quantum computing has been validated as a concept and has been accomplished experimentally on a very small scale. We are still far from a desktop or laptop quantum computer, but such a device seems to be on the horizon. Instead of the bit as the fundamental data unit as in classical computing, quantum machines would make use of the qubit. One qubit can represent a one or two plus any quantum superposition of these traditional states. So the information it can convey when arranged in a string is truly astronomical.
Thus one obvious benefit of molecular-scale electronics is the almost limitless amount of information it can hold and manipulate. Another, due to the property of entanglement, is that secure networks will become in principle beyond the gentle ministrations of the hacking community.
Leave a Reply
You must be logged in to post a comment.