Re-running a particular iteration of the montecarlo sims

Dear greywanderer,

greywanderer said:

I now want to run a more detailed simulation on iteration 77. Is there a way to do it? I tried using seed=77 but that does not seem to work.

Hmmm...I am not sure what simulator you are using nor what version of virtuoso/spectre/MMSIM you are using greywanderer. If I assume you are using a relatively recent version of Assembler to submit your Monte carlo simulation, method 2 of the Cadence On-line support article at URL:

https://support.cadence.com/apex/ArticleAttachmentPortal?id=a1O3w000009Eq0VEAS&pageName=ArticleContent

details how to reproduce a specific Monte Carlo simulation to do further analysis or debugging. You need to indicate not only the seed, but also the starting point and the statistical method as the article notes. I hope I fully understood your question and my assumptions about your simulation methodology/version are correct....and most importantly.... this helps provide some insight greywanderer.

Shawn

My assumption is that greywanderer wants to know how to do this in a spectre netlist as many questions previously have involved including Netlist-level analysis statements.

You wouldn't change the seed - in fact there's very little reason to change the seed unless you want to force different results. Instead you would set firstrun=77 numruns=1 on the Monte Carlo analysis. See "spectre -h montecarlo".

Regards,

Andrew

Yes, it makes sense that its not the seed. Because that would change the starting point of the random number sequence. The firstrun according to the spectre help is the "starting iteration number" and wasn't clear to me before that it is the iteration of the random number generator with a given seed. So it makes sense that with the same seed I should set firstrun=77 and numruns=1.

Shawn, Thanks for your response. This question was for netlist-level analysis statements as Andrew assumed.

Best regards,

Murthy

Dear greyeanderer,

> This question was for netlist-level

> analysis statements as Andrew

> assumed.

Rats! Sorry for not fully understanding your context! I hope my information did not confuse you or waste your time!

Nevertheless, I thank you Andrew for stepping in and providing the correct answer, and thank you greywanderer for letting me know I misinterpreted your question and use!

Shawn

Not at all..