Robert S Grimes’ recently published development story is a good case study in System Realization and SoC Realization. Grimes developed a quad laser control system for a high-power military application and the article he’s published in EETimes—Designing with core-based high-density FPGAs—illustrates many of the thought processes and tasks associated with System Realization and SoC Realization. Grimes could not use a microcontroller because the system’s timing requirements included the need to take 200 million samples/sec from an eight-channel pulse-detection-and-analysis subsystem that monitored the behavior of four Q-switched lasers. With just 5 nanoseconds between samples, a software-based, processor-centric approach to sampling wasn’t going to be sufficient so the sampling system clearly had to be implemented in hardware. However, other system requirements for control and monitoring clearly called for a software-based system to handle the expected complexity so this design clearly needed custom hardware, processors, and custom software.
Grimes’ article takes you through his thinking at each step in the System Realization process. He identifies what needs to be in software and what must be realized in hardware. He looks for appropriate hardware IP, develops some on his own and modifies some to meet the system’s needs. He uses open-source software where possible. In all, this article provides a small case study of precisely the sorts of tasks envisioned by the EDA360 concept as defined by System Realization and SoC Realization. If this project had been slated for volume production, it might have required silicon realization as well, but the military and experimental nature of the project favored an FPGA implementation path.
For more details, read the article using the above link. Note: This is a 6-page article but the page picker at the bottom of each page only goes to 5 for some reason, so you need to click at the end of the text on page 5 to get to page 6.
