Hi everyone!
In RTL synthesis, we can guide FSM encoding to optimize for power and area. I’m curious how this is handled in HLS… do we get any control over FSM encoding, or is it entirely managed by the tool?
Would love to hear how you all approach FSMs in HLS!
HLS tools automatically create Finite State Machine for the created Datapath.
Method of handling FSM’s might vary from tool to tool but most HLS tools provide different types of encoding such as one-hot, gray and binary encoding to reduce the power consumption of the design based on the number of states which is created by the tool.