Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

how clock skew can be beneficial

Status
Not open for further replies.

vikram789

Member level 3
Joined
Aug 1, 2008
Messages
59
Helped
2
Reputation
4
Reaction score
2
Trophy points
1,288
Activity points
1,671
hi, till now i have read that clock skew pose a problem for designers but now i have heard that it can also help by decreasing the clock period .. following two eq. might also help u to compreheand the idea so tht u can explain that to me....
1. T >= reg + path_{max} + S - (s_d - s_s)
2. (s_d - s_s) =< reg + path_{min} - H

* T is the clock period,
* reg is the source register's clock to Q delay,
* pathmax is the path with the longest delay from source to destination,
* S is the setup time of the destination register
* pathmin is the path with the shortest delay from source to destination,
* H is the hold time of the destination register,
* (sd − ss) represents the clock skew from the source to the destination registers,
* sd is the clock skew to the destination register, and
* ss is the clock skew to the source register.
 

First a correction: sd and ss are not skews, they are the insertion delays to those registers. Their difference is the skew.

Yes, skew can be beneficial in some cases. It is called "useful skew". The idea is that one register-to-register stage can have plenty of positive slack while the next stage has a setup violation (negative slack). So, what you could do is to reduce the insertion delay of the first stage's destination flop (ss) - this reduces its positive slack. But that destination flop is also the *next* stage's launching flop - and the reduction in insertion delay is sd for this *next* stage - and this helps increase the available time to solve its timing violation.

It is also sometimes called "time borrowing". One stage borrows time from the previous or next stage in the flow.


(NOTE: To interpret the setup and hold equations correctly it is absolutely essential that you understand that in equation 1 (setup), the ss is the insertion delay of the launching clock transition, and sd is the insertion delay to the capturing flip-flop of THE NEXT CLOCK TRANSITION (i.e.: T seconds later)
For equation 2 (hold) ss and sd refer to THE SAME CLOCK TRANSITION propagating to the two flip-flops.)
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top