Spectre Sims Do Not Update

E

El Andrew

Guest
Hello,

I'm using spectre through Cadence analog design environment 5.0.33. A
persistent problem has been that after running two simulations spectre
no longer updates changes to either variables, analysis settings, or
schematic values and just runs through the previous simulation again
(or perhaps just recalls results from the previous sim). It's not
until I quit and restart analog environment that an updated simulation
can be run (twice, until the same problem starts up again).

Looking in the simulation file I see that the necessary changes to the
netlist "input.scs" have been made, but it doesn't seem to matter as
an old netlist is being invoked to generate the results. I did notice
that in /.artist_states the "analyses" file has not been updated
however. Specifically, a sweep setting that I changed in analog
environment (and that was updated in "input.scs") was not changed in
the "analyses" file.

Hoping someone can be of help here, I haven't been able to get very
far with this very annoying problem.

Thanks,

El
 
El:
See previous post entitled "spectre interactive mode". We have
disabled interactive mode because of this problem with the following
setting:
envSetVal("spectre.envOpts" "controlMode" 'string "batch")
---
Erik

eleanorandrew@hotmail.com (El Andrew) wrote in message news:<e89378b2.0410042019.2c63e22a@posting.google.com>...
Hello,

I'm using spectre through Cadence analog design environment 5.0.33. A
persistent problem has been that after running two simulations spectre
no longer updates changes to either variables, analysis settings, or
schematic values and just runs through the previous simulation again
(or perhaps just recalls results from the previous sim). It's not
until I quit and restart analog environment that an updated simulation
can be run (twice, until the same problem starts up again).

Looking in the simulation file I see that the necessary changes to the
netlist "input.scs" have been made, but it doesn't seem to matter as
an old netlist is being invoked to generate the results. I did notice
that in /.artist_states the "analyses" file has not been updated
however. Specifically, a sweep setting that I changed in analog
environment (and that was updated in "input.scs") was not changed in
the "analyses" file.

Hoping someone can be of help here, I haven't been able to get very
far with this very annoying problem.

Thanks,

El
 
Hi Erik,

Thanks, but with some outside help, I have at last solved this nasty
little problem. It all boiled down to Cadence chocking on mouthfuls
of licensing utilities due to problems with the license pointer. Now,
it start up much faster, there's no latency in getting Spectre sims
underway, and, most importantly, I can run one session of Analog
Environment 'till the cows come home.

El

erikwanta@starband.net (Erik Wanta) wrote in message news:<84018314.0410060102.a1f9114@posting.google.com>...
El:
See previous post entitled "spectre interactive mode". We have
disabled interactive mode because of this problem with the following
setting:
envSetVal("spectre.envOpts" "controlMode" 'string "batch")
---
Erik

eleanorandrew@hotmail.com (El Andrew) wrote in message news:<e89378b2.0410042019.2c63e22a@posting.google.com>...
Hello,

I'm using spectre through Cadence analog design environment 5.0.33. A
persistent problem has been that after running two simulations spectre
no longer updates changes to either variables, analysis settings, or
schematic values and just runs through the previous simulation again
(or perhaps just recalls results from the previous sim). It's not
until I quit and restart analog environment that an updated simulation
can be run (twice, until the same problem starts up again).

Looking in the simulation file I see that the necessary changes to the
netlist "input.scs" have been made, but it doesn't seem to matter as
an old netlist is being invoked to generate the results. I did notice
that in /.artist_states the "analyses" file has not been updated
however. Specifically, a sweep setting that I changed in analog
environment (and that was updated in "input.scs") was not changed in
the "analyses" file.

Hoping someone can be of help here, I haven't been able to get very
far with this very annoying problem.

Thanks,

El
 
El:
Please explain what you did to resolve your licensing issue.
---
Erik

eleanorandrew@hotmail.com (El Andrew) wrote in message news:<e89378b2.0410071429.75d5483a@posting.google.com>...
Hi Erik,

Thanks, but with some outside help, I have at last solved this nasty
little problem. It all boiled down to Cadence chocking on mouthfuls
of licensing utilities due to problems with the license pointer. Now,
it start up much faster, there's no latency in getting Spectre sims
underway, and, most importantly, I can run one session of Analog
Environment 'till the cows come home.

El

erikwanta@starband.net (Erik Wanta) wrote in message news:<84018314.0410060102.a1f9114@posting.google.com>...
El:
See previous post entitled "spectre interactive mode". We have
disabled interactive mode because of this problem with the following
setting:
envSetVal("spectre.envOpts" "controlMode" 'string "batch")
---
Erik

eleanorandrew@hotmail.com (El Andrew) wrote in message news:<e89378b2.0410042019.2c63e22a@posting.google.com>...
Hello,

I'm using spectre through Cadence analog design environment 5.0.33. A
persistent problem has been that after running two simulations spectre
no longer updates changes to either variables, analysis settings, or
schematic values and just runs through the previous simulation again
(or perhaps just recalls results from the previous sim). It's not
until I quit and restart analog environment that an updated simulation
can be run (twice, until the same problem starts up again).

Looking in the simulation file I see that the necessary changes to the
netlist "input.scs" have been made, but it doesn't seem to matter as
an old netlist is being invoked to generate the results. I did notice
that in /.artist_states the "analyses" file has not been updated
however. Specifically, a sweep setting that I changed in analog
environment (and that was updated in "input.scs") was not changed in
the "analyses" file.

Hoping someone can be of help here, I haven't been able to get very
far with this very annoying problem.

Thanks,

El
 
Hi Erik,

All I did (was told to do actually) was to set the LM_LICENSE_FILE
with only the licenses I needed for that particular Cadence session.
Apparently the CDS_LIC_FILE pointer is not doing its job in keeping
license utilities under control. What a relief to finally make
changes and be able to simulate, I was on the verge of completely
switching over to ADS before this problem was solved.

Best,

El

El:
Please explain what you did to resolve your licensing issue.
---
Erik

eleanorandrew@hotmail.com (El Andrew) wrote in message news:<e89378b2.0410071429.75d5483a@posting.google.com>...
Hi Erik,

Thanks, but with some outside help, I have at last solved this nasty
little problem. It all boiled down to Cadence chocking on mouthfuls
of licensing utilities due to problems with the license pointer. Now,
it start up much faster, there's no latency in getting Spectre sims
underway, and, most importantly, I can run one session of Analog
Environment 'till the cows come home.

El

erikwanta@starband.net (Erik Wanta) wrote in message news:<84018314.0410060102.a1f9114@posting.google.com>...
El:
See previous post entitled "spectre interactive mode". We have
disabled interactive mode because of this problem with the following
setting:
envSetVal("spectre.envOpts" "controlMode" 'string "batch")
---
Erik

eleanorandrew@hotmail.com (El Andrew) wrote in message news:<e89378b2.0410042019.2c63e22a@posting.google.com>...
Hello,

I'm using spectre through Cadence analog design environment 5.0.33. A
persistent problem has been that after running two simulations spectre
no longer updates changes to either variables, analysis settings, or
schematic values and just runs through the previous simulation again
(or perhaps just recalls results from the previous sim). It's not
until I quit and restart analog environment that an updated simulation
can be run (twice, until the same problem starts up again).

Looking in the simulation file I see that the necessary changes to the
netlist "input.scs" have been made, but it doesn't seem to matter as
an old netlist is being invoked to generate the results. I did notice
that in /.artist_states the "analyses" file has not been updated
however. Specifically, a sweep setting that I changed in analog
environment (and that was updated in "input.scs") was not changed in
the "analyses" file.

Hoping someone can be of help here, I haven't been able to get very
far with this very annoying problem.

Thanks,

El
 

Welcome to EDABoard.com

Sponsor

Back
Top