OCV, AOCV, and POCV

Fig1: Derate Factor on Setup Analysis

         Fig2: Derate Factor on Hold Analysis

Issues in OCV:

 

Advance On-Chip Variation (AOCV):

                 Fig3: Bounding box for cell and net distance

Distance: If the distance increases, systematic variation will increase, and to mitigate the variation, we need to use a higher derate value. So along with the distance, the derate value increases.

Path depth: If the distance is fixed and path depth increases, systematic variation would be constant, but random variation would tend to cancel each other. Therefore, as path depth increases, the derate factor decreases.

Fig. 4: Path depth in the timing path

Cell type: The derate is based on the cell type, as an AND gate and an OR gate cannot exhibit the same variation. The derate value also varies with the drive strength of the cell; for example, AND2X2 and AND2X6 will have different drive derate values.

AOCV Analysis Mode:

AOCV supports two analysis modes:

  1. Clock only
  2. Clock and data

In clock-only mode, AOCV derate is applied only on the clock path, so it has reduced effort and a fast runtime. whereas in clock and data modes, AOCV derate is applied on full design. AOCV analysis requires an AOCV derate table for all the cells. Due to reduced timing pessimism in AOCV, it has been observed that huge time violations were fixed when we moved to clock-only AOCV derate mode, and the violations were further reduced when we moved to clock and data.

Setup 2 4 6
10 1.10 1.08 1.04
20 1.15 1.10 1.06
30 1.17 1.13 1.09

Fig5: AOCV 2D derate table example

Prime Time AOCV Flow:

Fig6:PrimeTime AOCV Analysis Flow                                                         

Issues in AOCV:

Parametric On-chip Variation (POCV):

POCV Analysis:

POCV Delay Calculation:

Delay of a cell = Nominal delay (µ) ± (C * Nominal delay) * N

Where C = POCV coefficient and

  N = Number of standard deviations

OR

Delay of cell = Nominal delay (µ) ± Variation

PrimeTime POCV analysis flow:

A flow of the primeTime tool for POCV derate analysis has shown in figure 8.

Fig 7: PrimeTime POCV analysis flow

 Comparison between POCV and AOCV:

AOCV POCV
Random Variation is modeled through the depth based derate and Systematic Variation is modeled through distance based derate. Random and Systematic Variations are modeled through a delay variation coefficient which is specific to each cell.
Less accurate correlation between GBA and PBA. More accurate correlation between GBA and PBA.
Transition Variation and cell check variation are not supported. Transition Variation and cell check variation supported.

Disclaimer:

The purpose of this blog is to spread knowledge and facilitate open discussions. The content shared is based on our research from diverse sources including the internet, books, and conversations with others. We strive for accuracy; however, we acknowledge the possibility of errors or omissions. We value your input and encourage you to share your comments, questions, and suggestions. We are open to clarifying or correcting any inaccuracies found within the blog. 

Please note that while we appreciate your contributions, we may not be able to implement all suggestions. The blog should not be considered a substitute for professional advice.