AFTERWORD

It is interesting to ask software people what they like about their work. Most say that intellectual stimulation is foremost. But if that is the case, then why wouldn’t proving mathematical theorems be just as stimulating? Surely that exercise requires the same type of intellectual activity? But theorem-proving lacks the feeling of creation that accompanies the writing of a program. Both are logical constructions in particular formal systems—but the theorem simply “is,” while the program (eventually!) “runs.”

Those who prefer hardware to software often point to the absence of constraints as the basis of their choice. They are free to create ab initio a machine that solves a problem optimally, rather than merely programming a suboptimal solution on an arbitrary machine. But to date, hardware has lacked the immediacy—instant gratification—of software.

This book has shown how the field-programmable gate array (FPGA) has brought hardware and software closer together. An FPGA would be virtually useless without extensive CAD support, including algorithms that represent some of the toughest challenges to software engineers and scientists. The act of programming an FPGA captures some of the immediacy of software design along with some of the freedom of hardware design.

From a systems perspective, FPGAs are often thought of as a replacement for “glue logic,” or for application-specific integrated circuits (ASICs), which have a longer design and fabrication cycle. But there is a potential for FPGAs to have a profound effect on how systems are designed at the fundamental level. The fact that an FPGA can be quickly reprogrammed in system means that one can design into a system a component whose functions change as the need arises. Like the microprocessor before it, the FPGA may replace hundreds of single-function components with one programmable part.

A system can be built with FPGAs and random-access memory (RAM) with instructions on how to configure the hardware, interleaved with other instructions on how to process data. Such a system is reminiscent of a living organism, in which the DNA codes the arrangement of basic building blocks needed to create the proteins that will make the organism function. This book points the way toward the next generation of systems with even higher levels of integration and functionality than ever before, with FPGA structures at the core.

CHARLES D. STORMON

President and Chief Scientist
Coherent Research Inc.
East Syracuse, New York