The Limits of Modern Computing
Moore’s law states that the maximum number of transistors that can be reliably fabricated on silicon doubles every 18 months. This has been the general trend for the past 50 years.
However, feature sizes are already in the tens of nanometers which is causing issues for semiconductor fabricators. Some of these problems include wet etching issues, resolution of projected designs, and the difficulty in getting more than 1 billion transistors perfect.
Other problems are more physical and bizarre, including quantum tunneling where electrons literally teleport from one conductor to the next (jumping across an insulating barrier) and result in a leakage current. This leakage current generates more heat which, in turn, increases leakage current resulting in thermal runaway.
Getting transistors even smaller is possible but whether those devices will be reliable enough to produce the latest processors is increasingly becoming an unlikely gamble. Even if 2nm transistors become a reality, after that point transistors will be made using individual atoms—feature size reduction beyond that is impossible.
It is, therefore, no surprise that scientists need to find alternative methods for computational methods that can meet the constantly growing demand for more powerful devices. Some possibilities include specialized devices such as photon computers which can solve complex algorithms such as the traveling salesman problem. Other solutions include the use of quantum computers to offload complex tasks from conventional computers such as encryption, database lookups, and path-finding algorithms.
One issue with conventional computers is that they have to be designed to accept any task. This comes at a cost to speed. A classic example is using a CPU in a computer to handle graphics and consequently having processing time taken away from programs. The ZX series of computers developed by Sinclair Research used this method—this meant that the user had to halt the display to maximize the processing power.
This problem has been overcome with the introduction of graphics cards which offload as much graphical processing as they can. The result has been graphics chips that can solve complex polynomial equations with powerful graphics features and more processing power for user programs.
This use of specialized hardware could be the key to improving computer performance in the future which is why genetic computing is highly sought after.
Some of the most complex processes known to man are those found in living cells. DNA is arguably the most complex.
DNA, also known as deoxyribonucleic acid, is the molecule that holds all the genetic instructions needed to make a living creature. DNA consists of four bases: A (adenine), T (thymine), G (guanine), and C (cytosine). Each base on the DNA strand potentially stores two bits of information.
The human genome consists of over 3 billion bases. It stands to reason, then, that each human cell potentially holds up to 6.4 billion bits (740MB) of information. Genetic computers, however, would not be limited to high-density data storage because DNA can also be manipulated to create structures in a process known as “DNA origami”, the nanoscale folding of DNA molecules to create arbitrary 2D and 3D structures by exploiting the interaction between the base pairs A, T, G, and C. Using DNA origami, a team of scientists even developed a DNA computer that can play tic-tac-toe.
But DNA devices on their own will not be sufficient for modern computing as there needs to be a method for bridging electrical components to organic processors. This is where a research team from Germany has made headway. They've taken on that problem with their research into gold-coated DNA nanowires.
Genetic material shows promise in the field of computing
The team, including Bezu Teschome and Artur Erbe of Helmholtz-Zentrum Dresden-Rossendorf, have created their gold nanowires using a series of complex steps, the first of which involves DNA origami. The structure is first designed as a 3D or 2D raster model which is then fed into a computer. The computer determines how the structure will fit together and then designs a DNA strand with specific base pairs at certain points. When this strand (along with other material) is mixed in solution, heated, and then cooled, it forms the desired design due to Watson-Crick base pairing.
DNA Origami can be used to create structures as shown here. Image courtesy of Thomas H. LaBean and Hao Yan via Michael Strong. [CC BY 2.5]
This nanotube, however, is not very conductive—which is a problem when the goal is to connect such genetic material to electrical components. Therefore, the team needed to increase the conductivity of the nanotubes which was done by using gold.
Special molecules were used to bind gold ions to the outer layer of the nanotube to (theoretically) make the structure conductive. But there was still a problem with the design. Unless the tube can be connected to electrodes for testing, the electrical conductivity cannot be tested.
To tackle this problem, the team used electron beam lithography to help electrically connect the strand to probes that have a tip width of just a few tens of nanometers. The structure was tested between room temperature and 4 Kelvin with a result of Ohmic behavior which demonstrates conductivity throughout the tube.
DNA nanotube structure. Image courtesy of Graham D. Hamblin via McGill University
The Genetic Advantage
Organic devices have many advantages over electrical devices that could truly revolutionize how data is processed and handled—and how technology interacts with living beings.
For example, genetic material can self-replicate which gives the possibility of self-replication and self-healing computers. Such an ability may, in turn, lead to more advanced robotics and intelligent systems which can self-program and replicate with no human interaction at all.
Genetic material also has the ability to be readily absorbed by biological entities such as human beings which could be the key to computational implants of the future. DNA nanowires could be used to connect to individual neurons in the brain and spinal cord which may be used for those who suffer paralysis and/or those who require artificial limbs.
The other major advantage of DNA systems is the independence from electrical energy. DNA systems are dependent on heat and phosphates (ATP) as their source of energy which could be paired with electronic devices perfectly. Electrical circuits generate heat as a byproduct, which is unwanted. But if a DNA co-processor was coupled in the same package as a semiconductor, then the waste heat could be used to power the DNA machine at no extra cost.
DNA nanowires could be used to connect neurons to electrical devices. Image courtesy of Zeiss [CC BY 2.0]
Manipulating genetic material and shaping it into almost any conceivable structure seems like the stuff of science fiction. While it is still too early to make claims about how genetic computers may be used and how they may change technology, it cannot be denied that they will change technology if they are implemented.
But such ideas may also be controversial. Use of genetic material and implantation may be a mere stepping stone away from genetic ideology and manipulation.