Cadence MonteCarlo using HSPICE models

[tstoll]

Hi,
I have to perfom a MonteCarlo simulation. I have only Spectre available, and the only models I have are Hspice models.
The standard simulation is working fine.
But I do not get the MonteCarlo to run.

It stops with that error: ERROR (SPECTRE-16011): mc1: Attempt to run Monte Carlo analysis with mismatch variations, but no mismatch variations were specified in the statistics block.

The model files I have got, include a Monte Carlo file from Hspice like:

.variation
.global_variation
C CA SCALE=0.030
C CB SCALE=0.010

......

Is there a way to make that run in Spectre?

Regards,
Tom

 

[dick_freebird]

Seems to say that if you want to run with mismatch, you have
to tell it what mismatches are valid. Seems fair. Either tell it
to run with only process variation, or make up some mismatch
data and put it where the simulator expects to find it.

There may be a parallel statistics-enabled models hierarchy
that you just aren't pointing to, in the design kit. Or, you may
have to create some based on other info (like using the mismatch
vs L, W, area charts to deduce Pelgrom coefficients and so on).

 

[tstoll]

So you think there is no way to get the original Hspice MonteCarlo model into spectre?
I only have hspice models available. Standard simulation is working.
If that is not possible, any idea if only the parameter/statistics block generated in spectre syntax than could solve the issue?