FPGAs – Driving Computing, Military, Medical, Electronics, Wireless and Many More Applications
“Adding FPGAs change the characteristics of a supercomputer, fundamentally. As the FPGA is reconfigurable one is able to design a non von Neumann network topology, specific to the algorithm with one cycle latency. The algorithm is programmed at a very fine-grain parallelism, a single instruction on each node, each node adapted to run its instruction. The architecture transforms sequential instruction scheduling into parallel packet switching.”
Another example is Intel where researchers are using the FPGA to build a simulator system by placing 1000 processors in forty FPGA units. The size, power consumption, and cost are far below what a traditional supercomputer architecture costs. More specifically, for this configuration a traditional architecture would cost $2M while the FPGA version cost $100K.
FPGAs (Field Programmable Gate Arrays) offer tremendous promise to the field of data acquisition, a key technology in the world of Virtual Implementation. Chips with increased A/D sampling rates are matched by the increased capabilities of FPGAs thus offering improved real-time performance. In this article the author outlines the advantages of applying an A/D converter to an FPGA for real-time processing.
FPGAs are also used in high-end imaging systems, such as medical imaging. Pixel Velocity offers an FPGA-based imaging system with very low NRE costs. It takes some effort to build up the infrastructure in an FPGA to support all the peripherals required, but once established, it becomes highly leveraged by letting the designer easily add new features. The other challenge is the programming aspect. Again, an infrastructure needs to be built that allows for additional modifications. Once in place, the designer reaps the benefits.
FPGAs are increasingly used in consumer electronics where components must be low-cost. Initially used as glue logic, FPGAs are now used for the core logic.
In the military, FPGAs find use in Software Defined Radios by combining both IF (intermediate frequency) and baseband processing into one processor.
In the communications area, cellular, broadband internet wireless, and data security are driving the use of FPGAs for the complex front end processing.
For those engineers trying to make the jump to FPGAs, here’s an interesting article that highlights difference between ASIC design and FPGA-based design. In ASIC design, the cost of a design turn drives verification and testing, while the FPGA allows the designer to drive toward a performing function without the associated cost of creating new masks.