In CPRE 558 I learned a lot about real time systems. We covered many scheduling algorithms. One thing we did not explore extensively but I considered a lot was the high costs of scheduling algorithms. Specifically the cost of context switching for a task to preempt and having an OS to calculate what tasks to go when.

To explore an option to mitigate the context and scheduler overhead my project partner and I decided to try to implement RMS scheduling in hardware on an FPGA. We were able to successfully implement it. In the future I would like to explore putting it with a processor and be able to better compare the implementation costs and performance.