cpe229_lec08 - CPE 229 Course Notes: Lecture 8 Copyright:...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
CPE 229 Course Notes: Lecture 8 Copyright: 2005 Bryan Mealy The Big FSM Picture The world progressed really nicely for billions of years without having the concept of finite state machines or many other such important matters. In regards to FSMs, the course of events that have led us to where we are today is this (a few details have been left out): Digital stuff (computers and things) started happening. The digital stuff required some circuitry to control it; FSMs were the logical option. The thought here is that maybe you could use a computer to control a computer, but, these were still the days where computers were actually expensive (and big). Discrete components such as gates and flip- flops were used to construct FSMs which could be used to control things and this was good. Integrated circuits started taking over. This meant that some of the control functions that had been relegated to FSMs were now being built into the various ICs. There were also ICs dedicated to controlling specific devices which were essentially required because devices were becoming complex and control requirements were also growing in complexity. Microcontrollers started becoming prevalent. This meant that the hardware being controlled by FSM could now be controlled in many circumstances by MCUs. This meant that hardware devices could now essentially be under software control (what drives the MCU) rather than hardware (what FSMs are constructed from). The upside of this software control is the flexibility in software (re-programmability). The downside is that using the MCU to control hardware requires processing time from the MCU or dedicating a MCU to the control task. As this becomes an intensely complex problem once you have one MCU in charge of controlling many tasks. The push here is to choose the smallest/cheapest MCU to get the job done. This means a less capable MCU is in charge or more tasks. This gets ugly. Programmable Logic Devices (PLDs) such as FPGAs and CPLDs hit the market and became BIG (in size at least). This means that you could use the PLD to handle logic functions required by your circuit. Since you may have already been using a PLD, it makes sense to use that PLD to implement a FSM while you’re at it. And what the heck, you could use the PLD to implement the entire MCU (known as a soft-core MCU). In other words, transferring control from the MCU to the FSM would not be too costly. In the end, MCUs are often limited by the number of pins they have to interface to the outside world. The pin count is generally related to the cost of the MCU also. Now days, MCUs can do a lot of tasks (generally at the same time, sort of), which is good. The downside of having MCUs do a lot of tasks is that the associated software becomes more complicated and error prone based on the number of tasks it is required to control. The type of errors associated with digital systems such as these are intermittent and hard to reproduce and repair. So the point here is that FSMs are not dead. There is still a use for them and they are still used quite often
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 13

cpe229_lec08 - CPE 229 Course Notes: Lecture 8 Copyright:...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online