Re: Timing optimization interview questions can anyone answe
1. Timing analysis is done first to determine if the design will work for the specified frequency. For the first level of optimization, we specify the initial operating frequency and optimize the design for it. After this optimization is done, the reports state the maximum delays , critical paths, etc.
2. Normally, we do not specify the required frequency in the first timing run. We optimize it for a lower frequency and then we optimize the design for power and area.
The design is then checked for functionality and then subject to further optimizations till the required clock speeds are obtained.
However, in ASIC design there is a trade-off between Speed - Power - Area. So if a design is optimized for speed only, it will not meet our area or power requirements. After a few iterations of optimizations, we arrive at an optimal solution which satisfies or nearly meets all our requirements.
3. If after optimization, the skews are not met, we further optimize using other algorithms such as register re-timing, pipelining, etc till we meet our skew requirements. In some cases we may relax our constraints.