Dynamic gate with conditional keeper for soft error rate reduction
First Claim
Patent Images
1. A dynamic gate comprising:
- a node;
a pull-down network to conditionally discharge the node LOW during an evaluation phase; and
a pull-up to charge the node HIGH by switching ON during the evaluation phase, wherein the pull-up switches ON only if the pull-down network does not conditionally discharge the node LOW during the evaluation phase.
1 Assignment
0 Petitions
Accused Products
Abstract
A dynamic logic gate with a conditional keeper, the conditional keeper comprising a pMOSFET pull-up that switches ON only after the dynamic logic gate completes an evaluation so as to avoid contention with the pull-down network. By sizing the conditional keeper to be stronger than the half-keeper, embodiments may realize a significant reduction in soft error rates that are latched.
21 Citations
13 Claims
-
1. A dynamic gate comprising:
-
a node;
a pull-down network to conditionally discharge the node LOW during an evaluation phase; and
a pull-up to charge the node HIGH by switching ON during the evaluation phase, wherein the pull-up switches ON only if the pull-down network does not conditionally discharge the node LOW during the evaluation phase. - View Dependent Claims (2, 3, 4)
-
-
5. A dynamic gate comprising:
-
an input port to receive a clock signal;
a ground rail;
a power rail;
a node;
a pull-down network to provide a conditional low impedance path between the node and the ground rail only if the clock signal is HIGH;
a delay element to provide a signal indicative of a delayed clock signal, the delayed clock signal lagging in phase with respect to the clock signal; and
a pull-up to provide a low impedance path between the node and the power rail only if the node and delayed clock signal are HIGH. - View Dependent Claims (6, 7, 8)
-
-
9. A dynamic gate comprising:
-
an input port to receive a clock signal;
a node having a node signal;
a pull-down network connected to the node to conditionally discharge the node HIGH only if the clock signal is HIGH;
a pull-up pMOSFET having a drain connected to the node and having a gate; and
a static logic gate having a first input port to receive the clock signal, a second input port connected to the node, an output port connected to the gate of the pull-up pMOSFET;
the static logic gate to provide a logic signal indicative of a delayed version of the clock signal, and to provide at its output port a NAND function of the logic signal and the node signal. - View Dependent Claims (10, 11)
-
-
12. A computer system comprising:
-
a bus;
a memory unit coupled to the bus; and
a microprocessor comprising;
a node;
a pull-down network to conditionally discharge the node LOW during an evaluation phase; and
a pull-up to charge the node HIGH by switching ON during the evaluation phase, wherein the pull-up switches ON only if the pull-down network does not conditionally discharge the node LOW during the evaluation phase. - View Dependent Claims (13)
-
Specification