1 of 1 people found this helpful
The file format may be identical for the .db and .gds files.
Maybe a quick way to test this would be to create a link with a .db extension, to the .gds file then run the job and see if it works properly. Or if the file is small enough you might simply copy it to a new name with .db extension (or just rename the gds file to .db file) and try the job.
I will try that and let you know the result
Hi again Chris
You are right, both files can work for each other.
I am trying to run Calibre xRC from command line. I have the .gds file and .src.net file and also pex.runset
I am editing this pex.runset by changing the .db extension into .gds
and it is working properly.After extraction, it creates many files including netlist , .pxi files
but this pxi file is empty.
Do you think why is that so?
By the way running calibre from command line is possible right?
I need output file for HSPICE.
Calibre can most definitely be run from the command line; that used to be the easiest way to run the newer features in PEX.
To use your runset in batch mode, try
calibre -gui -pex -runset <runset filename> -batch
To run PEX without Calibre Interactive, see the Calibre xRC User's Manual -- there are a ton of command line options depending on what you want to do.
HSPICE is certainly possible, but the exact SVRF command depends on the version of Calibre you are running. In the older versions, it was PEX Netlist Distributed or PEX Netlist Lumped, and you would give HSPICE as the format. I'm unsure of the newer format, but if you grep your __rules__ for "HSPICE" it should be there.
As for the empty .pxi file, there could be several reasons. The .pxi contains the instantiation of the parasitic models; if you have done a "Simple" (maybe called "ideal"
now) extraction, there are no parasitics to instantiate. (The "simple" (that is, no parasitics) model is for checking that the xRC output is logically identical to the input.) Another cause would be in a hierarchical extraction if the only shapes at the top level are ports; in that case, look for <cell_name>.pxi files. (This generalizes to if there was nothing at the top level to extract -- for example, if you excluded all the top-level nets somehow.) A third reason was that something was misconfigured -- look for warnings in the transcript. (Customer support is expert at sifting through all the messages, if you would like some assistance there.)
Thanks! I don't have any PEX experience and besides, it's just better when more people join in to help anyway I think.