Dears,
I'm currently facing the following problem:
When running PSS simulation using Spectre or APS (Multithreading Disabled), stabilization time simulation (and subsequent PSS iterations) runs through smoothly, leading to the following example plot:
However, just by enabling multithread APS simulation, the output signal swing is killed few nanoseconds after the period setup as stabilization time (100ns), as noted below:
This is the best I could do so far to track down to possible causes of such behavior. My design is mostly populated at this moment with analogLib ideal components, and (hidden-state free) verilog-a blocks that were bullet-proof tested (also with multithread APS simulation) on another server. Unfortunately I'm forced to move to this new server, since this is the only one with the PDK installation.
Virtuoso/Spectre version on faulty simulation server: Virtuoso 6.1.7 - 64b / Spectre 16.1.0 64bit
Virtuoso/Spectre version on successful multicore APS simulation server: Virtuoso 6.1.6 - 64b / Spectre 13.1.1 64bit
Thank you in advance for your support.