APPARATUS AND METHOD FOR LARGE HARDWARE FINITE STATE MACHINE WITH EMBEDDED EQUIVALENCE CLASSES

Patent №

US 7,180,328

Granted

2007-02-20

Filed 2004

Owner

SENSORY NETWORKS, INC.

Lab

AI components

1

hardware

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

10850978

A programmable finite state machine (FSM) includes, in part, a first address calculation logic block, a first lookup table, a second address calculation logic block, and a second lookup table. The first address calculation logic block generates an address for the first lookup table based on the received input symbol and the current state. The data stored in first look-up table at the generated address is used by the second address calculation logic block to compute an address for the second lookup table. Data stored in the second lookup table is the next state to which the FSM transitions. The programmable FSMs uses redundant information of the transition table to compress these transitions and thus requires a smaller memory while maintaining a high data throughput. The data in the first and second lookup tables are coded and supplied by a compiler. The FSM operation may optionally be pipelined.

AI classification

AI hardware0.89
Knowledge representation0.13
Evolutionary computation0.04
Natural language0.02
Machine learning0.01
Vision0.01
Planning0.00
Speech0.00

Ownership

SENSORY NETWORKS, INC.

assignment · 152690699

Assignors

GOULD, STEPHEN, PELTZER, ERNEST, BARRIE, ROBERT MATTHEW, FLANAGAN, MICHAEL, WILLIAMS, DARREN

On an employer assignment, the assignors are typically the inventors.

© 2026 NYSGPT2525 LLC