By Michael Ross
This article was published by ComputorEdge, issue #2407, 2006-02-17, as a feature article, in both their print edition (on pages 18 and 20) and their website.
As mankind's technological capabilities grow more formidable, and its scientists and engineers are increasingly able to borrow techniques from other fields, we are seeing more instances in which disparate technologies are being woven together — taking us in directions which we could scarcely have imagined not too long ago. For example, consider the following three, seemingly unrelated threads:
Firstly, if present trends continue, then sometime within the next 20 years — and possibly within the next 10 — the current technology for manufacturing microelectronics will reach its physical limit. One reason for this is the method by which each integrated circuit (IC) is created: The logical functioning of each IC is determined by the patterns of the many layers deposited on a wafer, using a process known as photolithography. The image of each desired pattern is projected onto the surface, which changes when exposed to the light, and thus able to be dissolved by a developing solution.
Short-wavelength ultraviolet light is the tool of choice for this work, because it is the shortest wavelength of visible light, at about 0.5 microns. Many chips now use patterns as small as 0.13 microns, which is unimaginably tiny. But because of the limitations of the ultraviolet light, the photolithographic microprocessors of the future can be made only so small.
Other restrictions include packing an increasing number of transistors into the same amount of space. In addition, more capable transistors require more energy. All of these factors, in turn, result in more heat being generated, which causes microchips and printed circuit boards to fail even faster.
The Software of Life
The second thread is the tremendous progress made by genetic researchers, microbiologists, and other scientists in analyzing and better understanding deoxyribonucleic acid (DNA). As most people know, DNA is the nucleic acid containing the genetic instructions that determine the development of all cellular forms of life, as well as the majority of viruses (but not the ones that end up inside your computer — just on the outside, when you eat at your computer).
Yet the genetic instructions of DNA take up remarkably little space, being composed of chains of nucleotides, which are themselves so small as to make the 0.13 microns of our computer chip patterns seem humongous in comparison. In fact, even though the polymer chain of a human chromosome could be stretched out to about 5 cm, the breadth of the polymer chain cannot even be detected by current methods of electron microscopy.
Thirdly, computer scientists are forging ahead in the field of self-assembling computer architectures. While much of the progress to date is of a theoretical nature, researchers can point to an ever-growing list of breakthroughs and promising ideas.
At this point, the majority of the work is divided between two domains: nanotechnology and DNA. In the case of the former, researchers are attempting to scale down electronic and mechanical devices to atomic dimensions. The favorite building block for such miniature contraptions are carbon nanotubes, as a result of their thermal and chemical stability.
Borrowing from Mother Nature
The three threads discussed above are now being combined to form the basis of what may eventually turn out to be the most revolutionary development in computer science: DNA-based self-assembling computer architectures. Not only would such a development greatly diminish the current microprocessor limitations of heat, energy, and size (among others), but it could also fundamentally alter the way that artificial computing is performed.
It is known that DNA can be utilized as a structural component for microcircuits, as follows: DNA is attached to the ends of silicon rods, which consequently determines how the rods are joined, based upon the sequences. After the strands join the rods together, they are metallized, which causes them to be electron conductive.
Admittedly, each component would have little computational power by itself, being so simple. But their value lies in their tiny size, because it allows for an enormous number of them to be combined into a functioning processor that is still a fraction of the size of the theoretical limit of our current microprocessors with identical computing abilities.
There are two primary architectural designs in use at this point. The first one, known as a decoupled array multi-processor (DAMP), has been demonstrated in a device that contains 4096 nodes, each of which has more than 265 million processors, for a total of 1 trillion processors. All of the processors communicate through a central control unit.
A contrasting design is the so-called "oracle" (not to be confused with the database manufacturer of the same name). Such a device is packed with innumerable question-answer pairs, and derives its functionality from the sheer number of pairs contained within, and not from the complexity of each pair — similar to the way that a DAMP's simplistic processors are combined in such numbers as to provide the needed computational muscle.
DNA and Computers
If and when researchers are first able to build complex microprocessors utilizing DNA, it certainly will not be the first time that DNA has played a critical role in the world of computers. According to many sources, the problems presented by genetic analysis have formed a significant and longtime motivation for researching problems in computer science, as well as other related fields.
For instance, much of the work done on string searching algorithms (finding occurrences of sequences of letters inside of much longer sequences) was motivated by the need to find and match the four types of DNA building blocks (abbreviated as A, T, C, and G) in genetic data. Researchers in the life sciences had a tremendous need to find specific sequences of nucleotides within larger sequences.
Databases, now an essential component in almost every large-scale computer application and e-commerce website, form another area of computer science that has greatly benefited from DNA-related research. Software developers and database programmers in the field of bioinformatics, have invented new techniques for storing DNA data in the specialized genomic databases that they create.
Regardless of where DNA-computing takes us in the future, it should be obvious to anyone with a working neural-based "processor" between their ears, that the end results will likely be astounding, and could open up whole new vistas of computing capabilities, as well as some serious ethical problems — especially if and when people begin experimenting with combining their own brains with artificially-constructed ones.
As there is theoretically no limit to the computational power achievable by a single DNA-based processor, it's quite possible that the human brain will eventually be unseated from its current position of Best Brain in the World. So keep an eye on those viruses… including the ones in your keyboard.
Copyright © 2005 Michael J. Ross. All rights reserved.