Relationship to Nanotechnology
 

 
If "nanotechnology" and "computer architecture" sound like unfamiliar jargon to you, we have created a metaphorical explanation that you can read before, or instead of, this page.
 
Nanotechnology is nanometer-scale technology, or technology that is one or more orders of magnitude smaller than what is used today. A nanometer is a billionth of a meter, a millionth of a millimeter, ten Angstroms, or the combined diameter of about ten carbon atoms. Nanoelectronics is the development of electronic devices within which the smallest features are between 1 and 10 nanometers wide. Nanotechnology research is currently focused on the construction of useful objects such as computer memory out of particular molecules, or arrangements of carbon atoms, or quantum dots, while semiconductor technology is miniaturizing from microns down to hundreds of nanometers, and perhaps ultimately to 10 nanometers.
 
This ability -- through molecular engineering, quantum dots, processes evolved from CMOS, etc -- is still under development, and a picture of the manufacturing techniques that will be used at the 1 to 10 nanometer scale is not yet fully formable. However, from progress to date, a number of independent researchers have begun to describe what is attainable from nanoelectronics and desirable from a computer architecture standpoint. These descriptions read like a description of the Cell Matrix™ architecture: a regular, homogeneous, three dimensional or two dimensional array of simple reconfigurable elements with local-only interconnections, fault tolerance, dynamic fault handling, and better than linear configuration times.
 
If Cell Matrix technology is combined with nanotechnology, it benefits both technologies. The construction of Cell Matrices™ lets nanotechnology enter the digital circuit and system market sooner, because it permits the construction of a single, simple physical structure that is then electronically differentiated into any desired digital circuit or system through software. Nanotechnology benefits Cell Matrix technology because it provides inexpensive manufacture of Cell Matrices that contain enough cells to be useful for solving the kinds of large, difficult problems that people would love to solve today but cannot with today's technology. As an example, we have designed high performance parallel searchers that trade materials for time: much faster solutions using a lot more materials. But to build the machines we have designed we need significantly less expensive materials with a unit size orders of magnitude smaller than what is available today, and to end up with a machine with, say, 1018 (a trillion trillion) components that takes up a reasonable amount of space and that can be configured or set up quickly, we need cells arrayed in three dimensions rather than just two.
 
The Cell Matrix architecture acts as a bridge between nanotechnology and the world of electronic components: nanotechnology can be used to build Cell Matrix hardware; Cell Matrices can then be used to implement electronic components, circuits, and systems, and this three step process is significantly easier than going straight from nanotechnology to arbitrary electronic components, circuits and systems.
 
Why is a bridge needed? In the distant future, it may not be needed: humans may figure out how to build any electronic component they need in an extremely compact representation at the atomic scale. In the meantime, the Cell Matrix architecture makes it possible to make steady progress toward atomic scale computing elements and to make useful and saleable products sooner.
 
For certain types of hardware and for certain classes of problems, the physical features of Cell Matrix hardware, such as low power requirements and ability to withstand manufacturing defects and later damage, will make the use of Cell Matrices preferable over literal physical translations of the component. And for other classes of problems such as Avogadro scale computers -- that is, computers that efficiently use on the order of 1023 components -- both the physical features such as the need for little external interconnection, and the functional features of the Cell Matrix architecture, such as fast configuration times and dynamic, self-configuring hardware, cause the Cell Matrix hardware to become an end in itself rather than a means, or bridge.