Computer-aided design and analysis of circuits and semiconductor – Nanotechnology related integrated circuit design
Reexamination Certificate
1998-10-09
2001-08-07
Smith, Matthew (Department: 2825)
Computer-aided design and analysis of circuits and semiconductor
Nanotechnology related integrated circuit design
Reexamination Certificate
active
06272667
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention is related to a gated clock design technique by the use of a computer, and more particularly related to a clock supplying circuit and an enable buffer cell, and a computer aided design system and method for clock gated logic circuits and design method thereof and an enable buffer cell.
2. Description of the Prior Art
In recent years, the integration and the chip size of LSIs have dramatically increased to require much power consumption. The gated clock design technique has been developed for the purpose of obtaining the logic circuit designs with low electric power consumption.
The procedure of modifying logic circuits by means of the gated clock design technique is called as “clock gating” in the following explanation. First, the gated clock design as mentioned above will be briefly explained.
FIG. 1
is a partial circuit diagram showing a clocked logic circuit which has not been clock gated. In the same figure, the reference symbols FF
0
,FF
1
, . . . FF
31
designate 32 D-type flip-flops in the form of a 32 bits data register in combination. The clock input port CK of the respective flip-flop is given the clock signal CLK at the same timing while the data input port D of the respective flip-flop is given data as calculated in the data processing circuit
11
in synchronism with the rising timing of the clock signal CLK.
It is known that the power consumption is increased when the clock signal supplied to the flip-flops rises up or falls down in the operation of the logic circuit of this kind. However, unnecessary power consumption occurs when there is no need for loading data to flip-flops, since the clock signal is continuously input to the flip-flops irrespective of existence of data to be loaded.
FIG. 2
is the circuit diagram of a logic circuit to which the clock gating technique is applied, in which an AND gate
12
is disposed as the gating circuit on the clock line as illustrated in FIG.
2
. The AND gate
12
is connected to the enable logic circuit (not shown) for controlling output of the clock signal and for outputting the enable signal E of a logic one or a logic zero to the AND gate
12
in synchronism with the clock signal CLK.
In
FIG. 2
, when the enable signals E is set a logic one, data is input to the flip-flops in synchronism with the clock signal CLK because the logic product of the enable signals E and the signal CLK is a logic one. On the other hand, when the enable signals E is set a logic zero, data is not input to the flip-flops irrespective of the clock signal CLK taking a logic one because of the logic product of the enable signals E and the signal CLK always remains of a logic zero. Accordingly, when there is no need for loading data to flip-flops, unnecessary power consumption can be avoided by outputting the enable signals E of a logic zero in synchronism with the clock signal CLK.
The power saving efficiency is largely depending upon the actual technique of how to generate the enable signals based upon which storage elements such as registers and memories are selectively given the clock signal. Prior to the present invention, the design of the enable signals has been conducted by manual processes or by CAD systems for automatically generating the enable signals from the control algorithm of the logic circuit.
However, in accordance with the prior art clock gating circuit designing technique, any proposals for generating appropriate enable signals have not been made from the view point of the power saving efficiency. For this reason, it is not confirmed that the enable signals as used are appropriate for the power saving efficiency so that it, is often the case that the clock gating circuit designing technique has been conducted with inefficient enable signals.
On the other hand, while the frequency of the clock signal as supplied to the circuits integrated within an LSI chip has been drastically increased, the electric power consumption is also proportionally increased along therewith resulting in difficult problems associated. More specifically speaking, as a result of the increased electric power consumption, the heat as generated tends to exceed the capacity of heat dissipation of the package of an LSI chip. On the other hand, the life time of a battery system for supplying power to the LSI is often shortened by the resultant high temperature.
It is possible to assume that the sources of the power consumption of an LSI chip are located in a data transmission system, a clock system, an I/O system and function modules. Among them, the electric power consumption of the clock system is particularly large as compared to other systems. It is therefore effective for saving electric power consumption to decrease the electric power consumption in the clock system.
The clock gating technique has been well known as an effective technique for reducing the electric power The of a clock system. In accordance with this prior art technique, data transfer between registers is performed, only when it is necessary, by halting and resuming clock supply by means of a gated clock supplying circuit.
FIG. 3
is a circuit diagram showing a clock supplying circuit for the clock supplying a gated clock signal in accordance with the prior art technique. Enable buffer cells
72
are inserted between a root driver cell
71
and the register (flip-flop)
73
in order to form a tree structure. The respective enable buffer cells
72
are controlled by enable signals E
1
to E
3
. Meanwhile, in this case, there is a branch through which the registers
73
are connected directly to the root driver cell
71
.
In accordance with this gated clock circuit technique, the clock supplying circuit is implemented with the buffer cells
72
each of which outputs the logic product or sum of the clock signal CLK and the control signal E
1
, E
2
or E
3
called an enable signal are used. In the case as illustrated in
FIG. 3
, each enable buffer cell
72
outputs the logic product.
Accordingly, for example, only when the enable signals E
1
is logic one, “1”, the clock signal from the root driver cell
71
is passed to the register
73
at the end of sub-trees through the enable buffer cells
72
. In the description, the clock signal as passed through the enable buffer cell is simply called as “the clock signal” while the clock signal as originally input to the root driver cell
71
is called as “the clock signal CLK”.
FIG. 4
is a circuit diagram showing the operation of the enable buffer cell
72
for outputting the logic product of two input signals as described above. Each enable buffer cell
72
receives the enable signals EN through one input terminal thereof and the clock signal through the other input terminal and supplies its output signal to the registers
73
. The clock signal is transferred through the buffer cells
72
only when the enable signals E
1
is logic one “1”, but not transferred through the buffer cells
72
when the enable signals E
1
is logic zero “1”. In the case that the enable buffer cell is designed for outputting the logic sum of two input signals, the clock signal is transferred through the buffer cell only when the enable signals E
1
is logic zero “0”, but not transferred through the buffer cell when the enable signals E
1
is logic one “1”.
As explained in the above, it is possible to dispense with unnecessary transmission of the clock signal by inserting the buffer cells of logic AND or OR receiving the enable signals and the clock signal together with a combinational circuit for generating the enable signals EN, and therefore to reduce the electric power consumption.
On the other hand, after generating appropriate enable signals, there might be further problems as follows. Namely, it is not yet confirmed whether or not the clock skew can be minimized in order to guarantee normal operation at a predetermined operating frequency while a number of different enable signals are used in the same clock system.
Conventionally, a sub-tree is formed with the buffer cells under contro
Kitahara Takeshi
Minami Fumihiro
Nishio Seiichi
Usami Kimiyoshi
Do Thuan
Foley & Lardner
Kabushiki Kaisha Toshiba
Smith Matthew
LandOfFree
Method and apparatus for clock gated logic circuits to... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for clock gated logic circuits to..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for clock gated logic circuits to... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2469290