The debate over whether FPGAs will replace standard cell ASICs has not fundamentally changed over the last 15 years. ASICs are still bigger, faster, and less expensive per device than FPGAs. In fact, the ratio of silicon logic efficiency and core performance between FPGAs and ASICs is still roughly unchanged.

Why is it then, that ASIC design starts are declining and FPGA revenues are growing faster than the overall semiconductor industry? Two reasons: first, the economics of building standard cell ASICs no longer make sense for most applications; and second, FPGAs are big enough and fast enough to address a majority of today’s application requirements.

As more advanced process technologies become available, both of these factors will only further accelerate the market shift to programmable logic.

Reprinted with permission from Embedded Computing Design/January 2005. Article © OpenSystems Publishing.