A pragmatic approach to systems on a chip

In electronics, there is an expectation that things will get smaller, more complex and cheaper all the time, While these expectations are often met by semiconductor suppliers, there are practical limitations to how far we can take things. As circuit geometries shrink, we can add more and more complexity. This fact leads people to believe that all things electronic should be put on a single chip. Systems on a chip, hyper-integration, whatever you call it, both the value and sensibility of increasing integration in semiconductor design must be addressed. This paper will introduce both technical and commercial aspects of making practical decisions in system partitioning as related to systems on a chip or in most cases, sub-systems on a chip.