Field-Programmable Gate Arrays:Features of FPGAs

Features of FPGAs

If we order semiconductor manufacturers to make mask-programmable gate arrays, we have to wait several weeks and pay twenty-thousand to hundreds of thousands of dollars. But with FPGAs, we can program FPGAs in minutes by ourselves and need to pay in the tens of dollars. But a FPGA can pack only about one-tenth the number of logic gates in a mask-programmable gate array because devices for user programmability, such as SRAMs, non-volatile memory, and anti-fuses, take up large areas. Mask- programmable gate arrays are faster by two orders of magnitude and far cheaper for large production volume. Thus, for debugging or verifying logic design that needs to be done quickly, FPGAs are used, and then mask-programmable gate arrays are used for large volume production after completing debugging or verification.

As seen already, FPGAs are classified into two types, depending on the types of devices used for programmability: rewritable FPGAs and non-rewritable FPGAs. Then, they are accordingly used for completely different purposes. Rewritable FPGAs, such as Xilinx’s based on SRAMs and Altera’s based on non-volatile memory, can be repeatedly rewritten in minutes. Non-rewritable FPGAs cannot be changed once programmed, but still have the advantages of realizing inexpensive logic chips with faster speed.

The area size of different devices for programmability also gives different advantages and disadvantages. A non-volatile memory cell is roughly four to five times larger than anti-fuse; a memory cell of SRAM is two times larger than a non-volatile memory cell. Anti-fuses are much smaller. So, because of smaller parasitic capacitance, FPGAs based on anti-fuses tend to be faster than those based on non-volatile memory or SRAMs.