From raidalzobaidi at yahoo.com  Wed Nov  2 01:08:20 2011
From: raidalzobaidi at yahoo.com (Raaed)
Date: Wed, 2 Nov 2011 02:08:20 -0700 (PDT)
Subject: [Hotspot] error: singular matrix in lupdcmp /3D modeling and
	the Grid Model
In-Reply-To: <CAPE+kwCFrtFNLtvZFLkDaMtGe5YfxtNWRBvQPdrn+K8ZZrxGgA@mail.gmail.com>
References: <mailman.0.1318950655.1607.hotspot@mail.cs.virginia.edu>
	<1319224489.68189.YahooMailNeo@web110012.mail.gq1.yahoo.com>
	<CAPE+kwBRNdtFkDMt6Bjz09Yh_dD1Dnpgz=k6vcZJqRB=piu+hw@mail.gmail.com>
	<1319660778.39913.YahooMailNeo@web110016.mail.gq1.yahoo.com>
	<CAPE+kwCFrtFNLtvZFLkDaMtGe5YfxtNWRBvQPdrn+K8ZZrxGgA@mail.gmail.com>
Message-ID: <1320224900.82720.YahooMailNeo@web110012.mail.gq1.yahoo.com>

Thanks Karthik, 
?
Although it sounds not very practical but what I start doing is repeating the thermal simulation from the beginning with the accumulated power trace from the simulator. I think this way it will work without any lose in the accuracy but on expense of bigger and bigger execution time. I think I am fine concerning this point but now I have something else to ask :-) I hope I am not flooding you and the Hotspot community by my questions :-(.
?
I am doing thermal simulation for a 3 layers stacked on top of each other and I've notice that the Grid model in Hotspot (assuming 128x128 is selected) is generating only one grid set of results (i.e. 128x128 matrix) I was expecting to get three sets of these,?one for each layer?! am I missing something here? please advice.
?
All best,
?
Ra'ed


________________________________
From: Karthik Sankaranarayanan <ks4kk at virginia.edu>
To: Raaed <raidalzobaidi at yahoo.com>
Cc: "hotspot at mail.cs.virginia.edu" <hotspot at mail.cs.virginia.edu>
Sent: Saturday, October 29, 2011 12:53 AM
Subject: Re: [Hotspot] error: singular matrix in lupdcmp


Hi Ra'ed,

That is surprising indeed! Please do send us your floorplan file and we will take a look.

As for your second question, the approach you use might not be fully accurate since the steady state temperature file only stores temperatures at a block granularity.By storing and resuming from it, you lose the spatial gradients within a block. Whether this loss of accuracy is significant or not is your judgment call. The alternative would be to write the entire grid model vector (grid cell temperatures of all layers) to a file and resume from it.? 

Thanks,
-karthik


On Wed, Oct 26, 2011 at 1:26 PM, Raaed <raidalzobaidi at yahoo.com> wrote:

Thanks karthik ... I doubled checked but I didn't find anything wrong with my floorplan file! any way the same file?work OK with the grid model :-) so I am using the grid model at the moment :D
>?
>I've another important question I need help with! I'm trying to integrate the Hotspot with the network simulator I'm using (Noxim) but it keep showing a linking problem? so I did a workaround?as follows: After executing number cycles (sampling time) in the network simulator I will store the power trace in a file and call the HOTSPOT as external exe file. Hotspot then will read the power trace and store the steady state temp in another file which in return going to be read by the network simulator and so on.... am I doing the correct thing or this will affect the internal states of the HOTSPOT in the subsequents execution (keeping in mind that I'm using the old temp as initial for the next run).
>?
>Need you advices?
>?
>Thanks.
>?
>Ra'ed
>?
>
>
>From: Karthik Sankaranarayanan <ks4kk at virginia.edu>
>To: Raaed <raidalzobaidi at yahoo.com>
>Cc: "hotspot at mail.cs.virginia.edu" <hotspot at mail.cs.virginia.edu>
>Sent: Friday, October 21, 2011 9:34 PM
>Subject: Re: [Hotspot] error: singular matrix in lupdcmp
>
>
>
>Hi Ra'ed,
>
>The singular matrix error typically means that the floorplan has some unaccounted empty spaces / overlaps. Could you please check to make sure there are no gaps/overlaps in your floorplan?
>
>Thanks,
>-karthik
>
>
>
>On Fri, Oct 21, 2011 at 12:14 PM, Raaed <raidalzobaidi at yahoo.com> wrote:
>
>Greetings, 
>>?
>>I am new to HotSpot so please accept my apology if I am asking a silly question!
>>?
>>I'm having this error message "error: singular matrix in lupdcmp" when I am trying to run the HotSpot on my processor floorplan!? The processor build using 65nm technology so the dimensions of the blocks are very small (chip area is 21mmx12mm and the smallest block measures 0.75mmx0.5mm). I realized if I multiply the dimensions by a factor of 100 then this message disappear and the program show me some results?! any suggestions?
>>?
>>Many thanks in advance.
>>?
>>Ra'ed
>>?
>>phd research student
>>Newcastle University
>>
>>
>>_______________________________________________
>>HotSpot mailing list
>>HotSpot at mail.cs.virginia.edu
>>http://www.cs.virginia.edu/mailman/listinfo/hotspot
>>
>>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111102/1305fe24/attachment-0001.html 

From wh6p at virginia.edu  Wed Nov  2 06:46:53 2011
From: wh6p at virginia.edu (Wei Huang)
Date: Wed, 2 Nov 2011 09:46:53 -0500
Subject: [Hotspot] error: singular matrix in lupdcmp /3D modeling and
 the Grid Model
In-Reply-To: <1320224900.82720.YahooMailNeo@web110012.mail.gq1.yahoo.com>
References: <mailman.0.1318950655.1607.hotspot@mail.cs.virginia.edu>
	<1319224489.68189.YahooMailNeo@web110012.mail.gq1.yahoo.com>
	<CAPE+kwBRNdtFkDMt6Bjz09Yh_dD1Dnpgz=k6vcZJqRB=piu+hw@mail.gmail.com>
	<1319660778.39913.YahooMailNeo@web110016.mail.gq1.yahoo.com>
	<CAPE+kwCFrtFNLtvZFLkDaMtGe5YfxtNWRBvQPdrn+K8ZZrxGgA@mail.gmail.com>
	<1320224900.82720.YahooMailNeo@web110012.mail.gq1.yahoo.com>
Message-ID: <CANYeXBoEtJ-=NH=KRsaBtbwMQsOn1QN1do=UPB-tnEGO-JjgBA@mail.gmail.com>

Hi Ra'ed,

How did you do the 3-layer simulation? Was it done with 3 layers specified
within the .lcf file? The grid model results of the 3D case should have
temperatures of all the grid cells in one file (including the thermal
interface materials between the 3D silicon layers).

-Wei

On Wed, Nov 2, 2011 at 4:08 AM, Raaed <raidalzobaidi at yahoo.com> wrote:

> Thanks Karthik,
>
> Although it sounds not very practical but what I start doing is repeating
> the thermal simulation from the beginning with the accumulated power
> trace from the simulator. I think this way it will work without any lose in
> the accuracy but on expense of bigger and bigger execution time. I think
> I am fine concerning this point but now I have something else to ask :-)
> I hope I am not flooding you and the Hotspot community by my questions :-(.
>
> I am doing thermal simulation for a 3 layers stacked on top of each other
> and I've notice that the Grid model in Hotspot (assuming 128x128 is
> selected) is generating only one grid set of results (i.e. 128x128
> matrix) I was expecting to get three sets of these, one for each layer?! am
> I missing something here? please advice.
>
> All best,
>
> Ra'ed
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111102/5a13e776/attachment.html 

From ks4kk at virginia.edu  Thu Nov  3 11:33:19 2011
From: ks4kk at virginia.edu (Karthik Sankaranarayanan)
Date: Thu, 3 Nov 2011 12:33:19 -0700
Subject: [Hotspot] error: singular matrix in lupdcmp /3D modeling and
 the Grid Model
In-Reply-To: <1320224900.82720.YahooMailNeo@web110012.mail.gq1.yahoo.com>
References: <mailman.0.1318950655.1607.hotspot@mail.cs.virginia.edu>
	<1319224489.68189.YahooMailNeo@web110012.mail.gq1.yahoo.com>
	<CAPE+kwBRNdtFkDMt6Bjz09Yh_dD1Dnpgz=k6vcZJqRB=piu+hw@mail.gmail.com>
	<1319660778.39913.YahooMailNeo@web110016.mail.gq1.yahoo.com>
	<CAPE+kwCFrtFNLtvZFLkDaMtGe5YfxtNWRBvQPdrn+K8ZZrxGgA@mail.gmail.com>
	<1320224900.82720.YahooMailNeo@web110012.mail.gq1.yahoo.com>
Message-ID: <CAPE+kwBJ3O-M=xwcmZmkxMi_uKzLqYt=-n3pDkoFQEsJhMP2NQ@mail.gmail.com>

Hi Ra'ed,

The -grid_steady option outputs only the internal temperatures of the top
layer (in a form readily usable by the grid_thermal_map.pl script). If you
want to output the temperature of the grid cells of all layers, you would
have to write your own function based upon the dump_top_layer_temp_grid
function.

All the best
-karthik

On Wed, Nov 2, 2011 at 2:08 AM, Raaed <raidalzobaidi at yahoo.com> wrote:

> Thanks Karthik,
>
> Although it sounds not very practical but what I start doing is repeating
> the thermal simulation from the beginning with the accumulated power
> trace from the simulator. I think this way it will work without any lose in
> the accuracy but on expense of bigger and bigger execution time. I think
> I am fine concerning this point but now I have something else to ask :-)
> I hope I am not flooding you and the Hotspot community by my questions :-(.
>
> I am doing thermal simulation for a 3 layers stacked on top of each other
> and I've notice that the Grid model in Hotspot (assuming 128x128 is
> selected) is generating only one grid set of results (i.e. 128x128
> matrix) I was expecting to get three sets of these, one for each layer?! am
> I missing something here? please advice.
>
> All best,
>
> Ra'ed
>
>  *From:* Karthik Sankaranarayanan <ks4kk at virginia.edu>
> *To:* Raaed <raidalzobaidi at yahoo.com>
> *Cc:* "hotspot at mail.cs.virginia.edu" <hotspot at mail.cs.virginia.edu>
> *Sent:* Saturday, October 29, 2011 12:53 AM
> *Subject:* Re: [Hotspot] error: singular matrix in lupdcmp
>
> Hi Ra'ed,
>
> That is surprising indeed! Please do send us your floorplan file and we
> will take a look.
>
> As for your second question, the approach you use might not be fully
> accurate since the steady state temperature file only stores temperatures
> at a block granularity. By storing and resuming from it, you lose the
> spatial gradients within a block. Whether this loss of accuracy is
> significant or not is your judgment call. The alternative would be to write
> the entire grid model vector (grid cell temperatures of all layers) to a
> file and resume from it.
>
> Thanks,
> -karthik
>
> On Wed, Oct 26, 2011 at 1:26 PM, Raaed <raidalzobaidi at yahoo.com> wrote:
>
>  Thanks karthik ... I doubled checked but I didn't find anything wrong
> with my floorplan file! any way the same file work OK with the grid model
> :-) so I am using the grid model at the moment :D
>
> I've another important question I need help with! I'm trying to integrate
> the Hotspot with the network simulator I'm using (Noxim) but it keep
> showing a linking problem? so I did a workaround as follows: Afterexecuting number cycles (sampling time) in the network simulator I will
> store the power trace in a file and call the HOTSPOT as external exefile. Hotspot then will read the power trace and store the steady state
> temp in another file which in return going to be read by the network
> simulator and so on.... am I doing the correct thing or this will affect
> the internal states of the HOTSPOT in the subsequents execution (keeping
> in mind that I'm using the old temp as initial for the next run).
>
> Need you advices?
>
> Thanks.
>
> Ra'ed
>
>
>  *From:* Karthik Sankaranarayanan <ks4kk at virginia.edu>
> *To:* Raaed <raidalzobaidi at yahoo.com>
> *Cc:* "hotspot at mail.cs.virginia.edu" <hotspot at mail.cs.virginia.edu>
> *Sent:* Friday, October 21, 2011 9:34 PM
> *Subject:* Re: [Hotspot] error: singular matrix in lupdcmp
>
> Hi Ra'ed,
>
> The singular matrix error typically means that the floorplan has some
> unaccounted empty spaces / overlaps. Could you please check to make sure
> there are no gaps/overlaps in your floorplan?
>
> Thanks,
> -karthik
>
>
> On Fri, Oct 21, 2011 at 12:14 PM, Raaed <raidalzobaidi at yahoo.com> wrote:
>
>  Greetings,
>
> I am new to HotSpot so please accept my apology if I am asking a silly
> question!
>
> I'm having this error message "error: singular matrix in lupdcmp" when I
> am trying to run the HotSpot on my processor floorplan!? The processor
> build using 65nm technology so the dimensions of the blocks are very
> small (chip area is 21mmx12mm and the smallest block measures
> 0.75mmx0.5mm). I realized if I multiply the dimensions by a factor of 100
> then this message disappear and the program show me some results?! any
> suggestions?
>
> Many thanks in advance.
>
> Ra'ed
>
> phd research student
> Newcastle University
>
>
> _______________________________________________
> HotSpot mailing list
> HotSpot at mail.cs.virginia.edu
> http://www.cs.virginia.edu/mailman/listinfo/hotspot
>
>
>
>
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111103/3baf36e5/attachment.html 

From bartosz.wojciechowski at pwr.wroc.pl  Mon Nov  7 03:32:01 2011
From: bartosz.wojciechowski at pwr.wroc.pl (Bartosz Wojciechowski)
Date: Mon, 07 Nov 2011 12:32:01 +0100
Subject: [Hotspot]  interpretation of r_convec and c_convec parameters
Message-ID: <1320665521.17585.23.camel@maly>

Hello all,

I'm trying to make a HotSpot model of my cpu package and
can't figure out what r_convec and c_convec parameters in hotspot.config
really are. I prepared a simple floorplan with 5 functional blocks.
Then, I prepared a very long power trace with total power of 86W
distributed over the floorplan so that power density is equal in every
FB. Shouldn't temperature converge to T_ambient + r_convec*P_total?
It doesnt't. The output is about 10K hotter.

Obviously, r_convec and c_convec depend on other parameters (like
heat_sink size, specific heat and such). A quick look at the code shows
that r_convec and c_convec are divided between nodes of RC network
corresponding to the heat sink. But it is unclear for me how to model my
specific heatsink. I measured/estimated total thermal resistance of the
package in my computer and the RC time constant and now I'd like to be
able to make transient simulations.

Any help and/or clarifications appreciated.

-- 
Regards,
BW


From wh6p at virginia.edu  Mon Nov  7 09:34:09 2011
From: wh6p at virginia.edu (Wei Huang)
Date: Mon, 7 Nov 2011 11:34:09 -0600
Subject: [Hotspot] interpretation of r_convec and c_convec parameters
In-Reply-To: <1320665521.17585.23.camel@maly>
References: <1320665521.17585.23.camel@maly>
Message-ID: <CANYeXBrN+QRn9sYuroNC_TU-bb+nGS98bo4J19e6Z2cY8tgrPw@mail.gmail.com>

Hi,

Please double check that your floorplan doesn't have empty (undefined area)
and/or overlapping areas.
If your floorplan is fine, please post the command line you used that
caused the problem (assuming you are using the latest version of HotSpot).

Thanks,

-Wei

On Mon, Nov 7, 2011 at 5:32 AM, Bartosz Wojciechowski <
bartosz.wojciechowski at pwr.wroc.pl> wrote:

> Hello all,
>
> I'm trying to make a HotSpot model of my cpu package and
> can't figure out what r_convec and c_convec parameters in hotspot.config
> really are. I prepared a simple floorplan with 5 functional blocks.
> Then, I prepared a very long power trace with total power of 86W
> distributed over the floorplan so that power density is equal in every
> FB. Shouldn't temperature converge to T_ambient + r_convec*P_total?
> It doesnt't. The output is about 10K hotter.
>
> Obviously, r_convec and c_convec depend on other parameters (like
> heat_sink size, specific heat and such). A quick look at the code shows
> that r_convec and c_convec are divided between nodes of RC network
> corresponding to the heat sink. But it is unclear for me how to model my
> specific heatsink. I measured/estimated total thermal resistance of the
> package in my computer and the RC time constant and now I'd like to be
> able to make transient simulations.
>
> Any help and/or clarifications appreciated.
>
> --
> Regards,
> BW
>
> _______________________________________________
> HotSpot mailing list
> HotSpot at mail.cs.virginia.edu
> http://www.cs.virginia.edu/mailman/listinfo/hotspot
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111107/2e5258da/attachment-0001.html 

From bartosz.wojciechowski at pwr.wroc.pl  Tue Nov  8 08:48:03 2011
From: bartosz.wojciechowski at pwr.wroc.pl (Bartosz Wojciechowski)
Date: Tue, 08 Nov 2011 17:48:03 +0100
Subject: [Hotspot] interpretation of r_convec and c_convec parameters
In-Reply-To: <CANYeXBrN+QRn9sYuroNC_TU-bb+nGS98bo4J19e6Z2cY8tgrPw@mail.gmail.com>
References: <1320665521.17585.23.camel@maly>
	<CANYeXBrN+QRn9sYuroNC_TU-bb+nGS98bo4J19e6Z2cY8tgrPw@mail.gmail.com>
Message-ID: <1320770883.15850.133.camel@maly>

Hi,

I must have somehow failed to explain my problem. HotSpot worked just
fine, and so did my floorplan ;-) The problem was understanding what
heat sink-related parameters meant. In HS-4.2 there were only: c_convec,
r_convec, s_sink and t_sink. Now, in 5.0 there are also k_sink and p
p_sink. So one of my questrions was: In reality, resistance and
capacitance depend on the other parameters. However in config I can give
them arbitrary values. What is the relation between these parameters?
Reference to a paper is fine :-)

The answer to the other question I managed to find for myself (but,
please, correct me If I'm wrong). To model transient behavior of my
processor package and heatsink, I need r_conv and c_conv of the heat
sink. But total thermal resistance (core-to-ambient) is a sum of r_conv
AND core-to-heat_sink resistance. I perfromed simulation. Then
calculated r_core_to_heatsink = (T_max - T_ambient)/P_total - r_conv.
Then found new r_conv = estimated_R_total - r_core_to_heatsink. From
there: c_conv = estimated_RC/r_conv. This is somewhat simplistic, but
the results seem to be in good agreement with temperature measurements
from my CPU.

Thanks for great software!
Regards,
BW

Dnia 2011-11-07, pon o godzinie 11:34 -0600, Wei Huang pisze:
> Hi,
> 
> 
> Please double check that your floorplan doesn't have empty (undefined
> area) and/or overlapping areas. 
> If your floorplan is fine, please post the command line you used that
> caused the problem (assuming you are using the latest version of
> HotSpot).
> 
> 
> Thanks,
> 
> 
> -Wei
> 
> On Mon, Nov 7, 2011 at 5:32 AM, Bartosz Wojciechowski
> <bartosz.wojciechowski at pwr.wroc.pl> wrote:
>         Hello all,
>         
>         I'm trying to make a HotSpot model of my cpu package and
>         can't figure out what r_convec and c_convec parameters in
>         hotspot.config
>         really are. I prepared a simple floorplan with 5 functional
>         blocks.
>         Then, I prepared a very long power trace with total power of
>         86W
>         distributed over the floorplan so that power density is equal
>         in every
>         FB. Shouldn't temperature converge to T_ambient +
>         r_convec*P_total?
>         It doesnt't. The output is about 10K hotter.
>         
>         Obviously, r_convec and c_convec depend on other parameters
>         (like
>         heat_sink size, specific heat and such). A quick look at the
>         code shows
>         that r_convec and c_convec are divided between nodes of RC
>         network
>         corresponding to the heat sink. But it is unclear for me how
>         to model my
>         specific heatsink. I measured/estimated total thermal
>         resistance of the
>         package in my computer and the RC time constant and now I'd
>         like to be
>         able to make transient simulations.
>         
>         Any help and/or clarifications appreciated.
>         
>         --
>         Regards,
>         BW
>         
>         _______________________________________________
>         HotSpot mailing list
>         HotSpot at mail.cs.virginia.edu
>         http://www.cs.virginia.edu/mailman/listinfo/hotspot 
> 
> 



From ks4kk at virginia.edu  Tue Nov  8 10:32:50 2011
From: ks4kk at virginia.edu (Karthik Sankaranarayanan)
Date: Tue, 8 Nov 2011 10:32:50 -0800
Subject: [Hotspot] interpretation of r_convec and c_convec parameters
In-Reply-To: <1320770883.15850.133.camel@maly>
References: <1320665521.17585.23.camel@maly>
	<CANYeXBrN+QRn9sYuroNC_TU-bb+nGS98bo4J19e6Z2cY8tgrPw@mail.gmail.com>
	<1320770883.15850.133.camel@maly>
Message-ID: <CAPE+kwCkQojrFBLyof0=DPKdz1ozD+D=6aDNRb1hp5dheKEE-w@mail.gmail.com>

Hi Bartosz,

Yes the method you outline below should give you  a very good approximation
of r_convec and c_convec. If you are interested in knowing how we
calculated the default values in the absence of a heatsink model, please
checkout Section 2.5.2 (pages 21-22) here:

http://www.cs.virginia.edu/~ks4kk/pubs/dissertation.pdf

Thanks,
-karthik


On Tue, Nov 8, 2011 at 8:48 AM, Bartosz Wojciechowski <
bartosz.wojciechowski at pwr.wroc.pl> wrote:

> Hi,
>
> I must have somehow failed to explain my problem. HotSpot worked just
> fine, and so did my floorplan ;-) The problem was understanding what
> heat sink-related parameters meant. In HS-4.2 there were only: c_convec,
> r_convec, s_sink and t_sink. Now, in 5.0 there are also k_sink and p
> p_sink. So one of my questrions was: In reality, resistance and
> capacitance depend on the other parameters. However in config I can give
> them arbitrary values. What is the relation between these parameters?
> Reference to a paper is fine :-)
>
> The answer to the other question I managed to find for myself (but,
> please, correct me If I'm wrong). To model transient behavior of my
> processor package and heatsink, I need r_conv and c_conv of the heat
> sink. But total thermal resistance (core-to-ambient) is a sum of r_conv
> AND core-to-heat_sink resistance. I perfromed simulation. Then
> calculated r_core_to_heatsink = (T_max - T_ambient)/P_total - r_conv.
> Then found new r_conv = estimated_R_total - r_core_to_heatsink. From
> there: c_conv = estimated_RC/r_conv. This is somewhat simplistic, but
> the results seem to be in good agreement with temperature measurements
> from my CPU.
>
> Thanks for great software!
> Regards,
> BW
>
> Dnia 2011-11-07, pon o godzinie 11:34 -0600, Wei Huang pisze:
> > Hi,
> >
> >
> > Please double check that your floorplan doesn't have empty (undefined
> > area) and/or overlapping areas.
> > If your floorplan is fine, please post the command line you used that
> > caused the problem (assuming you are using the latest version of
> > HotSpot).
> >
> >
> > Thanks,
> >
> >
> > -Wei
> >
> > On Mon, Nov 7, 2011 at 5:32 AM, Bartosz Wojciechowski
> > <bartosz.wojciechowski at pwr.wroc.pl> wrote:
> >         Hello all,
> >
> >         I'm trying to make a HotSpot model of my cpu package and
> >         can't figure out what r_convec and c_convec parameters in
> >         hotspot.config
> >         really are. I prepared a simple floorplan with 5 functional
> >         blocks.
> >         Then, I prepared a very long power trace with total power of
> >         86W
> >         distributed over the floorplan so that power density is equal
> >         in every
> >         FB. Shouldn't temperature converge to T_ambient +
> >         r_convec*P_total?
> >         It doesnt't. The output is about 10K hotter.
> >
> >         Obviously, r_convec and c_convec depend on other parameters
> >         (like
> >         heat_sink size, specific heat and such). A quick look at the
> >         code shows
> >         that r_convec and c_convec are divided between nodes of RC
> >         network
> >         corresponding to the heat sink. But it is unclear for me how
> >         to model my
> >         specific heatsink. I measured/estimated total thermal
> >         resistance of the
> >         package in my computer and the RC time constant and now I'd
> >         like to be
> >         able to make transient simulations.
> >
> >         Any help and/or clarifications appreciated.
> >
> >         --
> >         Regards,
> >         BW
> >
> >         _______________________________________________
> >         HotSpot mailing list
> >         HotSpot at mail.cs.virginia.edu
> >         http://www.cs.virginia.edu/mailman/listinfo/hotspot
> >
> >
>
>
> _______________________________________________
> HotSpot mailing list
> HotSpot at mail.cs.virginia.edu
> http://www.cs.virginia.edu/mailman/listinfo/hotspot
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111108/cac62dda/attachment.html 

From davizhao at gmail.com  Wed Nov  9 19:35:31 2011
From: davizhao at gmail.com (Dali Zhao)
Date: Wed, 9 Nov 2011 19:35:31 -0800
Subject: [Hotspot] rk4 returns temperature array filled with nan's
Message-ID: <CAE9BO967w7KeeQUKcM88fVnz1vZ7VN_q2jXB8ZDkjdPDOu4STA@mail.gmail.com>

Dear HotSpot community,

  I am running into an "nan" problem. I am using Hotspot on a Linux
machine. I tried to calculate the transient temperature
(compute_temp_grid). The rk4 function returns a temperature array
filled with nan's. The first 2 calls to rk4 function are fine. The
first few nan's appear in the 3rd call. Then they propagate and fill
the whole array with nan's. I am not familiar with numerical
algorithms and need some help.

  I searched the mail list archive and found a message back in 2004.
http://www.cs.virginia.edu/pipermail/hotspot/2004-May/000026.html
  The 2004 solution was to increase MAX_UNITS, which does not apply to
the current version.

  I am using the grid model to simulate a 3-d configuration. What
might cause the problem?

Thanks

-- 
Dali Zhao

From raidalzobaidi at yahoo.com  Mon Nov 14 03:35:45 2011
From: raidalzobaidi at yahoo.com (Raaed)
Date: Mon, 14 Nov 2011 03:35:45 -0800 (PST)
Subject: [Hotspot] Secondary heat transfer path
In-Reply-To: <CAPE+kwCFrtFNLtvZFLkDaMtGe5YfxtNWRBvQPdrn+K8ZZrxGgA@mail.gmail.com>
References: <mailman.0.1318950655.1607.hotspot@mail.cs.virginia.edu>
	<1319224489.68189.YahooMailNeo@web110012.mail.gq1.yahoo.com>
	<CAPE+kwBRNdtFkDMt6Bjz09Yh_dD1Dnpgz=k6vcZJqRB=piu+hw@mail.gmail.com>
	<1319660778.39913.YahooMailNeo@web110016.mail.gq1.yahoo.com>
	<CAPE+kwCFrtFNLtvZFLkDaMtGe5YfxtNWRBvQPdrn+K8ZZrxGgA@mail.gmail.com>
Message-ID: <1321270545.67185.YahooMailNeo@web110006.mail.gq1.yahoo.com>

Greetings,
?
I hope every one?in the HotSpot mailing list is fine.
?
Dear all, any one can help?me to find what materials (types)?and?their?thermal conductivity constant?modeled in HotSpot for the secondary heat transfer path, i.e., the following layers:
?
Underfill
C4 Pads
Chip carrier 
Solder balls
PCB
?
I used HotSpot and I am doing writting-up at the moment and I need to list these materials if possible? Also I wonder if there is a template to acknowledge the HotSpot community that you can provide?
?
All best, Ra'ed
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111114/de22d555/attachment.html 

From solidrepellent at gmail.com  Wed Nov 30 02:53:12 2011
From: solidrepellent at gmail.com (Solid Repel)
Date: Wed, 30 Nov 2011 12:53:12 +0200
Subject: [Hotspot] 3D steady state temperatures
Message-ID: <CAAYqiTuP9eX0wH2UTjB2G+Nu4r+uOKoDhAUdgoy=DQs7TVvZ4A@mail.gmail.com>

Hi all,

I am running a Hotspot simulation of a 3D stacked system with 3 layers. I
need to know the grid level steady state temperatures of those 3 layers. I
have two questions.

1) Is there a command line option to dump steady state temperatures of
different layers? If yes, then could you please let me know that.

2) I was trying to use the following command. Which layers steady state
temperatures (file 3D.grid.steady) the following command would give?

./hotspot -c hotspot.config -f 3DLayer_0.flp -p 3DLayer.ptrace -model_type
grid -grid_layer_file 3DNOC.lcf -steady_file 3DNOC.steady -grid_steady_file
3D.grid.steady

Thanks a lot for your time.

Best Regards,
Kamesh.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111130/f3e375ea/attachment.html 

From stasot at microlab.ntua.gr  Wed Nov 30 03:24:52 2011
From: stasot at microlab.ntua.gr (Efstathios Sotiriou - Xanthopoulos)
Date: Wed, 30 Nov 2011 13:24:52 +0200
Subject: [Hotspot] 3D steady state temperatures
In-Reply-To: <CAAYqiTuP9eX0wH2UTjB2G+Nu4r+uOKoDhAUdgoy=DQs7TVvZ4A@mail.gmail.com>
References: <CAAYqiTuP9eX0wH2UTjB2G+Nu4r+uOKoDhAUdgoy=DQs7TVvZ4A@mail.gmail.com>
Message-ID: <1322652292.26523.6.camel@CELINE>

Hello

1. If you want to have the solid state temperatures for each layer, then
you should modify the source code of hotspot, in order to make it dump
the temperatures for all the layers. For now, there is no command line
option. To be more precise, modify the dump_steady_temp_grid() function
inside temperature_grid.c file.

2. According to the source code, the temperatures of the top layer will
be written. Precisely, for this example, the file 3D.grid.steady will
give the first declared layer inside 3DNOC.lcf (usually, it is the layer
with number 0).

Kind regards,
Efstathios





On Wed, 2011-11-30 at 12:53 +0200, Solid Repel wrote:
> Hi all,
> 
> I am running a Hotspot simulation of a 3D stacked system with 3
> layers. I need to know the grid level steady state temperatures of
> those 3 layers. I have two questions.
> 
> 1) Is there a command line option to dump steady state temperatures of
> different layers? If yes, then could you please let me know that.
> 
> 2) I was trying to use the following command. Which layers steady
> state temperatures (file 3D.grid.steady) the following command would
> give?
> 
> ./hotspot -c hotspot.config -f 3DLayer_0.flp -p 3DLayer.ptrace
> -model_type grid -grid_layer_file 3DNOC.lcf -steady_file 3DNOC.steady
> -grid_steady_file 3D.grid.steady
> 
> Thanks a lot for your time. 
> 
> Best Regards,
> Kamesh.
> _______________________________________________
> HotSpot mailing list
> HotSpot at mail.cs.virginia.edu
> http://www.cs.virginia.edu/mailman/listinfo/hotspot