Molecular Computers

Unit 1 Discussion 2

Eric W. Wasiolek

Futuring and Innovation

Dr. Rhonda Johnson

Describe an innovation idea that is not possible today but will be available in the next 15 to 20 years.

Molecular Computers.

We are nearing the end of Moore's law with the current chip technology.  Moore's law states that the number of transistors in an integrated circuit (chip) per square inch will double every year.  This has held true since 1965, and has become part of the basis for the continued miniaturization of digital devices.  However, Moore's law is about to run into physical limitations of the current photolithography process of putting (printing) ever more transistors per square inch on the current chemical surfaces (silicon, gallium arsenide) that limit miniaturization.  Fundamentally new technologies at the molecular and atomic level will lead to the ultimate miniaturization.  I.e., you can't build a transistor that is any smaller than a set of atoms unless you attempt to build a computer at the quantum level.  

I am a subscriber to scientific American, and have read some articles in that magazine on molecular, bio-molecular, and quantum computing.  These technologies are clearly very much in their infancy and wouldn't see any maturation for 15 to 20 years.  Nano-computers, as come call them (nano being at the billionth of a meter size) are done at the molecular level.  

I would like to state a little more strongly how different molecular computers are from traditional computers.
A traditional computer is just a voltage processor.  A one is represented as a high voltage and a zero as a low voltage.  In a logic gate, for example an AND gate, if two high voltages come in then a high voltage comes out, otherwise a low voltage comes out.  In an OR gate is two low voltages go in a low voltage goes out, otherwise a  high voltage goes out.

In a molecular computer there are no voltages.  It is not a voltage processor.  There are no lines and gates etched on a chemical surface to create a circuit.  There are only molecules or an assemblage of molecules.  A one is represented as one state of the molecule (or configuration), a zero is represented as another state of the molecule.  This is not an electronic machine, it is a chemical machine.  Yes there is some electronics involved in chemicals, as it is through electron sharing that atoms combine.  The next generation of computers will not be built by computer scientists or electrical engineers, it will be built by chemists and physicists. Computer science will still obtain as the LOGIC of the molecules will be like the logic gates of a traditional computer and the ones and zeros that make up the computer instructions and data will at an abstract level be the same.  So programming will be similar.  At the physical leve the two computers are completely different.  Not only is this a breakthrough technology, it is so completely different that breakthrough is not strong enough a term.

What is not possible today and what doesn't exist today is a molecular computer.  By computer is meant an entire working computer that can process information and perform tasks.  This requires a molecular CPU and large molecular memory both of which don't exist yet.  The current research has created some basic logic gates out of molecular components.  These logic gates need to be connected into circuits, and these circuits need to form basic CPU components like a sum-er, complementer, etc.. to be able create CPU components like an arithmetic-logical unit (ALU).  

Identify and discuss two of the forces that define it and that may facilitate or reduce its likelihood of success.

A couple of forces that will define the molecular computer technology and facilitate or reduce its likelihood of success are the setting up of labs to work on molecular computers at universities, and funding of labs at universities and in industry to continue and improve the work.

The work in universities would be done primarily in physics and chemistry departments and their associated labs.  Although this may appear to be computer science or electrical engineering, the ability to hold bits in memory with a molecule, and to process those bits (which are states of the molecule) through molecular logic gates involves primarily physical and chemical research assuming a basic understanding of the logic involved in creating gates, logic components, ALUs (logic units) and entire computers that comes from electrical engineering and computer science.  So to some extent this may involve a cross pollination of disciplines (chemistry, physics, computer science, and electrical engineering), but the work would primarily be done in physics and chemistry assuming knowledge of how computer hardware is built.

Funding of this research in industry would also be helpful.  This could be done in research units of companies or in incubator labs funded by some startup source.  Any practical application of this technology awaits much more development and wouldn't be available for some 20 years.

Government programs to pursue this research at national labs would be helpful as well.  This would require a funding source which might need to come from a cabinet department or from congress.