Design alternatives for DSP solutions As a designer of Digital Signal Processing systems, you have a large number of choices to implement your solution. Each solution has its strengths and weaknesses. The purpose of this chapter is to introduce Xilinx Field Programmable Gate Array (FPGA) technology and help you understand how FPGAs can be used for DSP system implementation. In the following sections, you will find a comparison of the implementation of a simple DSP function in both Programmable DSP (pDSP) and Gate Array technology. A brief explanation of Gate array technology is followed by a description of Xilinx FPGA technology. If you are already familiar with Gate array technology, you might want to skip to the last section which highlights Xilinx FPGA features. Programmable DSPs The most common vehicle for implementation of a DSP design is the programmable DSP or pDSP. The pDSP is an off-the-shelf part that is essentially a microprocessor tuned to DSP applications. pDSPs are highly flexible because you can program them again and again using a familiar high level language like C. They allow fast design iterations and reduce time to market. Typically a pDSP contains several functional units to process the signal stream. The designer encodes the algorithm into a program which is executed by the pDSP and is limited to a theoretical maximum data rate based on the speed and the number of multiplier/ accumulators in the device. Applications which require several computations must be broken up into a sequential stream of computations. For example, an 8-tap FIR filter requires 8 fpgadsp