Electrical computers: arithmetic processing and calculating – Electrical digital calculating computer – Particular function performed
Reexamination Certificate
2000-05-05
2002-09-10
Ngo, Chuong Dinh (Department: 2124)
Electrical computers: arithmetic processing and calculating
Electrical digital calculating computer
Particular function performed
C708S232000, C326S041000
Reexamination Certificate
active
06449628
ABSTRACT:
BRIEF DESCRIPTION OF THE INVENTION
This invention relates generally to logic circuits that are used to perform computationally intensive tasks. More particularly, this invention relates to a hybrid logic circuit that is useful for computationally intensive tasks, such as wireless communications.
BACKGROUND OF THE INVENTION
Programmable logic devices are widely used in the electronics industry. Conventional programmable logic devices typically comprise a general-purpose logic array and general-purpose interconnect. The typical logic array in general-purpose Field Programmable Gate Arrays (FPGAs) is designed to accommodate some random/control logic functionality and some structured/datapath logic functionality. Using FPGAs can be very inefficient when the programmed logic function comprises many datapath functions. Inefficient device utilization and performance degradation result when general-purpose random/control logic and routing resources are used for structured datapath functions. The more general-purpose the device architecture, the greater the inefficiency and performance degradation incurred by structured datapath functions.
In view of the foregoing, it would be highly desirable to provide a programmable architecture that improved the performance of structured datapath functions and improved device utilization efficiency.
SUMMARY OF THE INVENTION
A programmable datapath arithmetic array includes resources that are data buses connected to a matrix of data arithmetic units including fixed function units and programmable function units. In an exemplary embodiment, the programmable datapath arithmetic array includes only fixed function units. In another exemplary embodiment, the programmable datapath arithmetic array includes only programmable function units. Bidirectional interconnect is positioned between the data buses and the matrix of data arithmetic units to facilitate dynamic reconfiguration and operability of the programmable datapath arithmetic array.
Unlike field programmable gate arrays (FPGAs), the programmable datapath arithmetic array includes two organized logic resources, namely, datapath slices and datapath structures. In an exemplary embodiment, the programmable datapath arithmetic array comprises an array of datapath slices. For example, when the programmable datapath arithmetic array is a 1×N array, N represents the number of datapath slices. A datapath slice comprises an array of datapath structures. In an exemplary embodiment, if a datapath slice is a 1×M array, M represents the number of datapath structures. A datapath structure comprises an array of bit-slice blocks. In an exemplary embodiment, if a datapath structure is a 1×L array, L represents the number of bit-slice blocks. Bit-slice blocks comprise building blocks of the programmable datapath arithmetic array. A bit-slice block comprises a bit-specific portion and a common-control portion. In an exemplary embodiment, the programmable datapath arithmetic array comprises two types of datapath structures, namely fixed and re-programmable datapath structures. Fixed datapath structures implement a limited set of functions whereas re-programmable datapath structures implement a relatively larger set of functions.
The programmable datapath arithmetic array includes dedicated routing resources. Further, in a preferred embodiment, the programmable datapath arithmetic array comprises coarse-grained logic. In an exemplary embodiment, the programmable datapath arithmetic array is designed to facilitate and accelerate datapath functions. Examples of datapath functions are counters, incrementers, decrementers, shifters, scalers, adders, subtractors, accumulators, and decumulators. In an exemplary embodiment, datapath functions exhibit a uniformity of structure across all bits.
REFERENCES:
patent: 4580215 (1986-04-01), Morton
patent: 4935734 (1990-06-01), Austin
patent: 5231588 (1993-07-01), Agrawal et al.
patent: 5329470 (1994-07-01), Sample et al.
patent: 5398198 (1995-03-01), Mahant-Shetti et al.
patent: 5796623 (1998-08-01), Butts et al.
patent: 5943242 (1999-08-01), Vorbach et al.
patent: 5946219 (1999-08-01), Mason et al.
patent: WO 98/21725 (1998-05-01), None
Morphics Technology Inc.
Ngo Chuong Dinh
LandOfFree
Apparatus and method for programmable datapath arithmetic... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus and method for programmable datapath arithmetic..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus and method for programmable datapath arithmetic... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2869171