Interlocked pipelined CMOS

Details Interlocked pipelined CMOS Interlocked pipelined CMOS

1998-11-20

2001-01-30

Heckler, Thomas M. (Department: 2787)

Electrical computers and digital processing systems: support

Synchronization of clock or timing signals, data, or pulses

Reexamination Certificate

active

06182233

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention generally relates to Complementary Metal Oxide Semiconductor (CMOS) Integrated Circuits (ICs) and, more particularly, to an Interlocked Pipelined CMOS (IPCMOS) family of logic circuits.
2. Background Description
Circuit techniques are key to implementing the interlocked pipelined approach. In order for the approach to be viable, the circuit techniques must be extremely fast, simple, reliable , and very efficient. The general concepts of handshaking or interlocking and pipelining and asynchronous self-timing are not new and have all been proposed in a variety of forms for systems.
The Muller C-element described by Carver Mead and Lynn Conway in Chapter 7 of
Introduction to VLSI Systems
, published by Addison-Wesley Publishing Company Inc. (1980), and shown in
FIG. 18
is a bistable device whose output becomes a “1” only after all of its inputs are “1” and whose output becomes a “0” only after all of its inputs are “0”. The major problem with this circuit, aside from speed issues for large fan-ins, is that all its inputs must overlap for both the “0” and “1” cases or there will be no change in the output state. For complex interconnections of circuit blocks, commonly called macros, that have different speeds and interconnections this may not be the case and the circuit will not perform the intended function.
Another self-timed element and a two-cycle signaling scheme described by Ivan E. Sutherland in “Micropipelines”,
Communications of the ACM
, vol. 32, Number 6, pp. 720-738, June 1989, and in the book by Mead and Conway is given in FIG.
19
. In this scheme, the elements pass data to each other using Request and Acknowledge lines to control the sequence in an asynchronous fashion as shown in the wave forms. Aside from not dealing with how to solve the problem of multiple macros being interconnected, this approach is slow since the Acknowledge signal as shown in the wave forms is initiated when the output of the macro reaches a stable state. This means that after data is received, the macro must do its function before sending an acknowledgment which inserts additional delay in the cycle time.
Control logic with completion information embedded within data signals is used in a self-timed divider described by T. E. Williams et al. in “A Zero-Overhead Self-Timed 160 ns 54-b CMOS Divider”,
IEEE J. Solid State Circuits
, vol. 26, pp. 1651-1661, November 1991. In this design, local completion detectors and handshaking between fully asynchronous macros is used. The completion information is embedded in the data throughout the design by using a pair of wires for each bit. The individual completion signals from the bits in a data path are then combined using a C-element similar to the one described above. This approach needs dual wires for each data signal plus it faces the limitations of the C-element already mentioned above.
Fully pipelined architectures have been used to obtain very fast cycle times in SRAMs, as described by T. I. Chappel et al. in “A 2-ns Cycle, 3.8-ns Access 512-Kb CMOS ECL SRAM with a Fully Pipelined Architecture”,
IEEE J Solid-State Circuits
, vol. 26, pp. 1577-1585, November 1991. This approach works well in a very regular environment such as one finds in memory, but is extremely difficult or impossible to implement in a more general environment without some kind of handshaking to guarantee reliable operation.
SUMMARY OF THE INVENTION
It is therefore an object of the present invention to provide a form of logic circuit that results in extremely high performance pipelined operation.
It is another object of the invention to provide logic circuits which guarantees error free operation where variations in timing are compensated for automatically by the circuits.
It is a further object of the invention to provide a standard interface that makes it possible to interface different macro types easily.
According to the invention, there is provided a circuit technique, here referred to as Interlocked Pipelined CMOS (IPCMOS), that results in extremely high performance and reliable operation. It uses interlocking in both the forward and reverse directions. This “handshaking” guarantees error free timing and makes it possible to eliminate the need for a global clock at the macro level. Timing signals are generated locally at the macro level from the handshaking signals between macros. This makes it possible for the local circuits to deal with global timing variations caused by power supply noise, ACLV (across chip linewidth variations), and parameter variations. The macros operate in a pipelined mode with data advancing automatically from macro to macro with the timing controlled by the local handshaking signals. This pipelined operation results in an extremely fast cycle time. Another feature of IPCMOS is that the data inputs to a macro are only sampled when the data is in a valid state. making the concept of a standard macro interface possible. With this standard interface, different logic types such as static and dynamic circuits can be easily interconnected and the concept of reusable macros becomes a reality.


REFERENCES:
patent: 4855958 (1989-08-01), Ikeda
patent: 5027355 (1991-06-01), Stoica
patent: 5300831 (1994-04-01), Pham et al.
patent: 5305463 (1994-04-01), Fant et al.
patent: 5416362 (1995-05-01), Byers et al.
patent: 6034912 (2000-03-01), Isomura et al.
T. Williams, et al., “A Zero-Overhead Self-Timed 160-ns 54-b CMOS Divider”, IEEE Journal of Solid-State Circuits, vol. 26, No. 11, Nov. 1991.
T. Chappell et al., A 2-ns Cycle 3.8-ns Access 512-kb CMOS ECL SRAM with a Fully Pipelined Architecture, IEEE Journal of Solid-State Circuits, vol. 26, No. 11, Nov. 1991.
I. Sutherland, “Micropipelines”, Communications of the ACM, vol. 32, No. 6, Jun. 1989.
S. Schuster, et al., “On-Chip Test Circuitry for a 2-ns Cycle, 512-kb CMOS ECL Sram”, IEEE Journal of Solid-State Circuits, vol. 27, No. 7, Jul. 1992.
T. Werner, “Asynchronous Processor Survey”, IEEE Nov. 1997.
C. Seitz, “Sysetm Timing”, Introduction to VLSI Systems, Carver Mead—Lynn Conway, pp. 218-262.

Affiliated with

Also associated with

Rating

Interlocked pipelined CMOS does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Interlocked pipelined CMOS, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Interlocked pipelined CMOS will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2484635