Browse Prior Art Database

Isolation of Microcode Function Using System Identification Disclosure Number: IPCOM000029684D
Original Publication Date: 2004-Jul-08
Included in the Prior Art Database: 2004-Jul-08
Document File: 2 page(s) / 46K

Publishing Venue



Described is the isolation of microcode function via System Identification (SI) techniques.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Isolation of Microcode Function Using System Identification

Internal models of the microcode function, each individual paradigms, are constructed in parallel under a multithreaded operating system such as OS/2*. The internal models use the responses (outputs) of the microcode in question to known commands (inputs). How well each model compares to the system is measured and used to rank the models against one another as well as to direct evolutionary changes in the individual models themselves. The best internal model could then be used to describe the function of the microcode in question without actually having to read that microcode line-by-line.

       The first step in this SI approach to deriving internal models of microcode is to define all of the input and output states. Knowledge of the number of input and output states is critical, as is a description of those states. Input and output states could be linear, piecewise linear, nonlinear, discrete, Boolean, time-dependent, etc.

       Once the input and output states are known, the construction of internal models of the microcode can begin. Example models (paradigms) which could be considered in parallel with one another are as follows.

       In Boolean logic, each state has two permissible values, either a 0 or a 1. Operators include AND, OR, NOT, NAND, NOR. Boolean controllers exist, but they are limited in scope due to the binary nature of each input or output state. For any Boolean microcode, we would quantify it with a "truth table," showing the values of input states and the resulting values of the output states.

       Fuzzy logic goes beyond Boolean logic by permitting state values between the limits of 0 and 1. One fuzzy-logic chip performs fuzzy logic processing directly in hardware. Another fuzzy logic chip has a proportional-integral-derivative (PID) controller with fuzzy logic-based adaptive tuning. It is proposed that fuzzy microcode...