Could you please share your synthesized netlist and the CMP macro placement solutions so that we can use them to reproduce your ISPD 2023 paper numbers?

[bodhi91]

I think it is difficult to reproduce the ISPD 2023 numbers without the synthesized netlist and the CMP macro placement solution. So, could you please share these inputs so that we can run your script and reproduce your numbers?

[sakundu]

We have contacted relevant EDA company representatives to understand whether this is possible. Unfortunately, our current license agreement does not allow us to share these files. (The actual synthesized netlist is “work product” that directly enables benchmarking; benchmarking of commercial EDA tools is forbidden.)

As should be clear, we used Innovus and Genus version 21.1 for our experiments. If you follow the flow given in Figure 2 of the ISPD-2023 paper, you should be able to reproduce all results and reported numbers. Even with different tool versions, our experiments show very similar results (please see the discussion of our checking with versions 19.1, 20.1 and 21.1 in FAQ #14).

Bottom line: Please follow the steps given in the ISPD-2023 paper and use our scripts to reproduce the numbers reported in the ISPD-2023 paper. Please report any problems here, and we would be happy to assist you. (If EDA company policy changes permit public upload and sharing of synthesized netlists alongside the scripts that produced them, we will upload them here for you and everyone else to access.)