摘要:
A method for analyzing a gated clock design in which a disabling clock gating transition prevents an output transition from occurring, assuring that no clock glitching occurs. Delays and slews are computed so that the arrival time computation that includes clock and gate signal delays are computed at the output, providing tests which ensure that no glitch situation occurs. The delays and slews are computed using a static timing analysis, which includes situations such as a late and early arriving gate clock signals. The invention may be used in any static timing analysis test to ensure that a first transition on one input of a circuit prevents the propagation of a second transition on another input of the circuit.
摘要:
A method and structure of clock optimization including creating an initial placement of clock feeding circuits according to clock signal requirements; identifying clusters of the clock feeding circuits, wherein each cluster includes a distinct clock signal supply device to which each clock feeding circuit within the cluster is connected; changing pin connections between the clock feeding circuits and clock signal supply devices to switch selected ones of the clock feeding circuits to different clusters to reduce lengths of wires between the clock feeding circuits and the clock signal supply devices within each cluster; and adjusting positions of the clock feeding circuits within design constraints to further reduce the lengths of the wires.
摘要:
A method of distributed timing analysis for a network which has been partitioned into at least two partitions, with each partition being assigned to a separate timing analysis process which communicates with the other processes is provided. A depth first search is performed by the process assigned to the partition in which the node for which timing information is desired is located. When the depth first search encounters a node for which a timing value is required which is located in another partition, a request is immediately sent to the process corresponding to that partition for the required timing information. When the request for timing information from the other partition is answered, the associated node is placed in the propagation queue. Also, as a node is processed, successor nodes which have had their predecessors processed are added to the propagation queue. The nodes in the propagation queue are processed and timing values are computed. When there are no nodes in the propagation queue, global loop detection is performed. Additionally, incremental timing updates are performed when a change is made in the network.
摘要:
A method for synthesizing a logic circuit that is driven by a clock signal, and that has a plurality of clock domains each having a plurality of clock sinks. A semiconductor substrate is provided. All of the plurality of clock sinks of one clock domain are placed into at least one cluster of clock sinks on the semiconductor substrate. A clock sink-density of each cluster of clock sinks is approximately equal to or greater than a clock sink density of an integrated circuit. A first portion of the plurality of clock sinks of a domain have a higher sink density than a second portion of the plurality of clock sinks of the same domain. The first portion has a subregion and the second portion has a subregion. The subregion of the first portion is adjacent to the region of the second portion.
摘要:
A method of analyzing timing differences for common path pessimism removal for a circuit containing a locked loop which controls at least two legs of a clock tree is provided. The method comprises the steps of computing the early and late mode delays for the locked loop; computing the early and late mode delays for delay segments in the circuit; identifying delay segments in the feedback path which control the locked loop; and adjusting the early and late mode delays for the delay segments in the feedback path based upon the type of feedback control used in the circuit.
摘要:
A method of analyzing timing differences between arrival times of distinct signals propagating through a circuit comprising: (i) identifying a first beginning point for a first data path over which data arrival times are propagated to a data endpoint; (ii) identifying a second beginning point for a second data path to the data endpoint, and for a clock path to a clock endpoint, the endpoints constituting an endpoint pair; (iii) identifying at least one common point at which the second data path and the clock path diverge; (iv) propagating data arrival times to the data endpoint from the first and second beginning points, and tagging the data arrival times propagated along the second data path with one of the common points; (v) propagating clock arrival times to the clock endpoint from the second beginning point along the clock path, and tagging the clock arrival times with one of the common points, the data and clock arrival times constituting clock/data arrival time pairs at the endpoint pair; (vi) determining the slack at the endpoint pair for the time pairs and identifying a worst time pair causing the worst slack determination; and (vii) recomputing the slack at the endpoint pair for the worst time pair, if the data arrival time in the worst time pair is tagged with the same common point, by eliminating slack caused by data and clock arrival time propagation on a shared portion of the second data path and the clock path between the second beginning point and the same common point.
摘要:
A system for optimizing a logic network including expressing the logic network as an original graph having vertices, edges which connect the vertices and which represent connections in the logic network, and inversion markings for representing inverters in the logic network; determining a fundamental cycle(s) in the original graph; sorting the determined fundamental cycle(s) according to its parity; forming a final graph by processing the fundamental cycle(s) so as to optimize inverter placement therein while maintaining the parity thereof; comparing the inversion markings of the original and final graphs to determine a set of transformation locations in the logic network; and re-configuring the logic network in accordance with the determined transformation locations.