Definition of clock latency (clock insertion hold off): In sequential types, each timing route is activated by a clock signal that originates from a supply. The flops getting triggered through the clock sign are called sinks for the clock. Usually, clock latency (or clock insertion delay) is defined as being the quantity of time taken through the clock signal in traveling from its supply into the sinks. Clock latency includes of two components - clock resource latency and clock community latency.
Supply latency of clock (Resource insertion delay): Resource latency is defined as being the time taken through the clock sign in traversing from clock resource (may be pLL,oscillator or various other source) to the clock definition point. It is also referred to as source insertion hold off. It could be utilized to design off-chip clock latency when clock supply is just not part of the chip itself. Community latency of clock (Community insertion hold off): Network latency is defined as the time taken through the clock signal in traversing from clock definition point into the sinks in the clock. So, every single sink on the clock provides a unique community latency. If we speak about the clock, it'll have:Optimum network latency: Optimum of all of the network latencies,Minimal community latency: Least of many of the community latencies,Common network latency: Typical of all the community latencies.
Whole clock latency is supplied given that the sum of resource latency and community latency. To paraphrase, total clock latency in a point is given as follows:Clock latency = Supply latency + Network latency It truly is typically said that to get a sturdy clock tree, ‘sum of source latency and network latency for all sinks of the clock need to be equal’. If that is the scenario, the clock tree is said being balanced as which means that each of the registers are receiving clock for the exact time; i.e., clock skew is zero.