From nedahpg at gmail.com  Thu Mar  1 21:08:59 2012
From: nedahpg at gmail.com (Neda Hassanpour)
Date: Fri, 2 Mar 2012 08:38:59 +0330
Subject: [Hotspot] question about HotSpot 3D stacking feature
In-Reply-To: <CAPa2X=TcmBW5=qcXQMCS3mAEqwE2VboY92wLUGz6mRM_iD7POw@mail.gmail.com>
References: <CAPa2X=TcmBW5=qcXQMCS3mAEqwE2VboY92wLUGz6mRM_iD7POw@mail.gmail.com>
Message-ID: <CAPa2X=SOajsRSyWF7D6ZJHN55R9WuaKqHwcX8JB49dJwZ60UHg@mail.gmail.com>

Hi,



I am new in working with HotSpot thermal model. I want to use HotSpot in 3D
NoC architecture as a sensor. The chip I?m working with has 4 layers and
there are 16 (4*4) processing cores in each layer. In fact I have a 3D mesh
topology (4*4*4) in the chip and I want to calculate each core?s
temperature using HotSpot .



I have downloaded HotSpot 5.2 and read the How To page. I have also checked
the files and some parts of the software code but still have some questions
about how to use 3D stacking capability of HotSpot which I would be really
thankful if anyone helps me.



1. There are some example files like ev6.flp or gcc.ptrace in the folder of
software. Are these files referring to a single processor system? As I
mentioned before, I have 64 core processors in a chip and I want to
calculate their temperature. Other parts like different levels of cache are
not related to my work. How should I prepare suitable inputs (for ptrace
and flp file) in this situation? I did not find any place to give HotSpot
the number of cores in each layer.



2. Although in the examples in the How To page has been referred to
gcc.ptrace.orig and ev6.flp.orig files, I can?t find them in the software
folder. My other question is about gcc.grid.steady file. According to the
manual, this file?s content relates to steady state temperature in grid
model but file includes nearly 4000 lines. What does each line of this file
represent?



3. If all the chip layers were the same (for example all have 2D mesh
topology and 16 cores), can use the same flp file for all of them?



4. In a flp file, using for each layer, should I consider each processing
core as a unit? (Like what has been done in example.flp )



5. Is it necessary to run any other commands before running the ones below,
to use 3D stacking feature of HotSot?

*hotspot -c hotspot.config -f <some_random_file> -p example.ptrace \*

*           -o example.ttrace -model_type grid \*

*           -grid_layer_file example.lcf*



6. How should I generate the initial values in the example.ttrace file, for
the first run of above commands ? Can I fill the file with ambient
temperature for each element at the beginning?



7. In lcf file, should I consider the heat sink as a layer too? For
instance in my case study with 4 layers of process elements, should I
define 5 layers in lcf file?



Thanks in advance

Neda
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120302/65203cc4/attachment.html 

From wh6p at virginia.edu  Fri Mar  2 10:16:50 2012
From: wh6p at virginia.edu (Wei Huang)
Date: Fri, 2 Mar 2012 12:16:50 -0600
Subject: [Hotspot] (no subject)
Message-ID: <CANYeXBqa8RkUYuSqQ5GADCG8R1JVR-_jK6ykgiKpdozm4Wbi7Q@mail.gmail.com>

Hi Neda,

Please see my responses below with leading >>>:

-------- Original Message --------
Subject:        [Hotspot] question about HotSpot 3D stacking feature
Date:   Fri, 2 Mar 2012 08:38:59 +0330
From:   Neda Hassanpour <nedahpg at gmail.com>
To:     hotspot at cs.virginia.edu



Hi,

I am new in working with HotSpot thermal model. I want to use HotSpot in
3D NoC architecture as a sensor. The chip I?m working with has 4 layers
and there are 16 (4*4) processing cores in each layer. In fact I have a
3D mesh topology (4*4*4) in the chip and I want to calculate each core?s
temperature using HotSpot .

I have downloaded HotSpot 5.2 and read the How To page. I have also
checked the files and some parts of the software code but still have
some questions about how to use 3D stacking capability of HotSpot which
I would be really thankful if anyone helps me.

1. There are some example files like ev6.flp or gcc.ptrace in the folder
of software. Are these files referring to a single processor system? As
I mentioned before, I have 64 core processors in a chip and I want to
calculate their temperature. Other parts like different levels of cache
are not related to my work. How should I prepare suitable inputs (for
ptrace and flp file) in this situation? I did not find any place to give
HotSpot the number of cores in each layer.

>>> you'll have to have your own floorplan files for each layer (or if all
the silicon layers
use the same floorplan, one single floorplan file is enough). Once that is
done, you can specify
which flp file to use for each layer in the lcf file.

2. Although in the examples in the How To page has been referred to
gcc.ptrace.orig and ev6.flp.orig files, I can?t find them in the
software folder. My other question is about gcc.grid.steady file.
According to the manual, this file?s content relates to steady state
temperature in grid model but file includes nearly 4000 lines. What does
each line of this file represent?

>>> The *.orig files were in previous releases. If you need them, please
let us know.
As for gcc.grid.steady file, each line corresponds to a grid cell and its
steady-state temperature.

3. If all the chip layers were the same (for example all have 2D mesh
topology and 16 cores), can use the same flp file for all of them?

>>> Yes you can do that in lcf file.

4. In a flp file, using for each layer, should I consider each
processing core as a unit? (Like what has been done in example.flp )

>>> It really depends on what level of thermal details you want to model.
Sometimes, you can treat a core as a unit, sometimes, you may want to use
sub-core architecture blocks as units.

5. Is it necessary to run any other commands before running the ones
below, to use 3D stacking feature of HotSot?

/hotspot -c hotspot.config -f <some_random_file> -p example.ptrace \/

/           -o example.ttrace -model_type grid \/

/           -grid_layer_file example.lcf/

>>> You'll probably want to run a steady-state simulation first and set the
steady-state temperatures
as your initial temperatures. This will save a lot simulation time since
the warm-up phase from ambient
temperature can be very long due to the big thermal mass of heatsink etc.
>>> Also, we don't explicitly support dumping transient grid temperature
results into a file (that usually ends
up with huge files). If you prefer to do so, you can use existing dump data
functions as templates to create a
separate dump transient grid data function,

6. How should I generate the initial values in the example.ttrace file,
for the first run of above commands ? Can I fill the file with ambient
temperature for each element at the beginning?

>>> See 5.

7. In lcf file, should I consider the heat sink as a layer too? For
instance in my case study with 4 layers of process elements, should I
define 5 layers in lcf file?

>>> Heatsink and heatspreader (and other package layers if you choose to
model the secondary heat transfer
path in hotspot.config) are default layers, so there is no need to
explicitly specify those in the lcf file. All you need
is the silicon and TIM layers.

Hope this helps.

-Wei
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120302/50ad3cc2/attachment-0001.html 

From nedahpg at gmail.com  Sat Mar  3 20:52:40 2012
From: nedahpg at gmail.com (Neda Hassanpour)
Date: Sun, 4 Mar 2012 08:22:40 +0330
Subject: [Hotspot] (no subject)
In-Reply-To: <CANYeXBqa8RkUYuSqQ5GADCG8R1JVR-_jK6ykgiKpdozm4Wbi7Q@mail.gmail.com>
References: <CANYeXBqa8RkUYuSqQ5GADCG8R1JVR-_jK6ykgiKpdozm4Wbi7Q@mail.gmail.com>
Message-ID: <CAPa2X=TQ6VsU+-kWfYffOCfbfp-JTOWoPi=iNuiS1tn-b3yRuw@mail.gmail.com>

Hi,


Thank you for your responses, they were really helpful. The main reason I
want to work with HotSpot is taking thermal capacities into account for
calculating cores temperature in a 3D NoC architecture. For now I?m working
with a (4*4*4) 3D architecture chip.  I still have some questions and would
be grateful if you help me this time too.


1. I understood that in gcc.grid.steady file, each line corresponds to a
grid cell and its steady-state temperature. So, is it true that the number
of lines in this file is determined by the size of grid which is defined in
the hotspot.config file before? (in this example this size is 64*64)



2. According to your response to my fifth question, I should run a
steady-state simulation first and set the steady-state temperature as my
initial temperature. So, what I think I should do now, Is running the
commands below, generating the gcc.grid.steady file, and using it for 3D
simulation as a list of initial temperature, is this true?



       hotspot -c hotspot.config -f ev6.flp -p gcc.ptrace \

      -steady_file gcc.steady -model_type grid \

       -grid_steady_file gcc.grid.steady



3. In my work I don?t need sub-core architectural blocks thermal details
and the only thing that is important to me is the cores temperature. So, I
think I should consider each core as a unit in my work. If I want to do so
for my chip (the one with 4 layers and 16 cores in each one), how does it
affect on grid size? In order to generate a single temperature for each
core, should I define the grid size as 4*4? Unfortunately the concept of
grid is still a little vague to me.



4. In the lcf file, is the layer with number 0, the nearest one to the heat
sink or the farthest one?



Thank you

On Fri, Mar 2, 2012 at 21:46, Wei Huang <wh6p at virginia.edu> wrote:

> Hi Neda,
>
> Please see my responses below with leading >>>:
>
> -------- Original Message --------
> Subject:        [Hotspot] question about HotSpot 3D stacking feature
> Date:   Fri, 2 Mar 2012 08:38:59 +0330
> From:   Neda Hassanpour <nedahpg at gmail.com>
> To:     hotspot at cs.virginia.edu
>
>
>
> Hi,
>
> I am new in working with HotSpot thermal model. I want to use HotSpot in
> 3D NoC architecture as a sensor. The chip I?m working with has 4 layers
> and there are 16 (4*4) processing cores in each layer. In fact I have a
> 3D mesh topology (4*4*4) in the chip and I want to calculate each core?s
> temperature using HotSpot .
>
> I have downloaded HotSpot 5.2 and read the How To page. I have also
> checked the files and some parts of the software code but still have
> some questions about how to use 3D stacking capability of HotSpot which
> I would be really thankful if anyone helps me.
>
> 1. There are some example files like ev6.flp or gcc.ptrace in the folder
> of software. Are these files referring to a single processor system? As
> I mentioned before, I have 64 core processors in a chip and I want to
> calculate their temperature. Other parts like different levels of cache
> are not related to my work. How should I prepare suitable inputs (for
> ptrace and flp file) in this situation? I did not find any place to give
> HotSpot the number of cores in each layer.
>
> >>> you'll have to have your own floorplan files for each layer (or if all
> the silicon layers
> use the same floorplan, one single floorplan file is enough). Once that is
> done, you can specify
> which flp file to use for each layer in the lcf file.
>
> 2. Although in the examples in the How To page has been referred to
> gcc.ptrace.orig and ev6.flp.orig files, I can?t find them in the
> software folder. My other question is about gcc.grid.steady file.
> According to the manual, this file?s content relates to steady state
> temperature in grid model but file includes nearly 4000 lines. What does
> each line of this file represent?
>
> >>> The *.orig files were in previous releases. If you need them, please
> let us know.
> As for gcc.grid.steady file, each line corresponds to a grid cell and its
> steady-state temperature.
>
> 3. If all the chip layers were the same (for example all have 2D mesh
> topology and 16 cores), can use the same flp file for all of them?
>
> >>> Yes you can do that in lcf file.
>
> 4. In a flp file, using for each layer, should I consider each
> processing core as a unit? (Like what has been done in example.flp )
>
> >>> It really depends on what level of thermal details you want to model.
> Sometimes, you can treat a core as a unit, sometimes, you may want to use
> sub-core architecture blocks as units.
>
> 5. Is it necessary to run any other commands before running the ones
> below, to use 3D stacking feature of HotSot?
>
> /hotspot -c hotspot.config -f <some_random_file> -p example.ptrace \/
>
> /           -o example.ttrace -model_type grid \/
>
> /           -grid_layer_file example.lcf/
>
> >>> You'll probably want to run a steady-state simulation first and set
> the steady-state temperatures
> as your initial temperatures. This will save a lot simulation time since
> the warm-up phase from ambient
> temperature can be very long due to the big thermal mass of heatsink etc.
> >>> Also, we don't explicitly support dumping transient grid temperature
> results into a file (that usually ends
> up with huge files). If you prefer to do so, you can use existing dump
> data functions as templates to create a
> separate dump transient grid data function,
>
> 6. How should I generate the initial values in the example.ttrace file,
> for the first run of above commands ? Can I fill the file with ambient
> temperature for each element at the beginning?
>
> >>> See 5.
>
> 7. In lcf file, should I consider the heat sink as a layer too? For
> instance in my case study with 4 layers of process elements, should I
> define 5 layers in lcf file?
>
> >>> Heatsink and heatspreader (and other package layers if you choose to
> model the secondary heat transfer
> path in hotspot.config) are default layers, so there is no need to
> explicitly specify those in the lcf file. All you need
> is the silicon and TIM layers.
>
> Hope this helps.
>
> -Wei
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120304/ae456e6a/attachment.html 

From wh6p at virginia.edu  Sun Mar  4 08:43:54 2012
From: wh6p at virginia.edu (Wei Huang)
Date: Sun, 4 Mar 2012 10:43:54 -0600
Subject: [Hotspot] (no subject)
In-Reply-To: <CAPa2X=TQ6VsU+-kWfYffOCfbfp-JTOWoPi=iNuiS1tn-b3yRuw@mail.gmail.com>
References: <CANYeXBqa8RkUYuSqQ5GADCG8R1JVR-_jK6ykgiKpdozm4Wbi7Q@mail.gmail.com>
	<CAPa2X=TQ6VsU+-kWfYffOCfbfp-JTOWoPi=iNuiS1tn-b3yRuw@mail.gmail.com>
Message-ID: <CANYeXBpm3Ca=jUCuUtXAyBu50S6_Sz-6Xc7Bynx1_WrKEFXKWg@mail.gmail.com>

>
>
> 1. I understood that in gcc.grid.steady file, each line corresponds to a
> grid cell and its steady-state temperature. So, is it true that the number
> of lines in this file is determined by the size of grid which is defined in
> the hotspot.config file before? (in this example this size is 64*64)
>

>>> yes.

>
>
> 2. According to your response to my fifth question, I should run a
> steady-state simulation first and set the steady-state temperature as my
> initial temperature. So, what I think I should do now, Is running the
> commands below, generating the gcc.grid.steady file, and using it for 3D
> simulation as a list of initial temperature, is this true?
>
>
>
>        hotspot -c hotspot.config -f ev6.flp -p gcc.ptrace \
>
>       -steady_file gcc.steady -model_type grid \
>
       -grid_steady_file gcc.grid.steady
>
>
>
 >>> Looks good to me.

3. In my work I don?t need sub-core architectural blocks thermal details
> and the only thing that is important to me is the cores temperature. So, I
> think I should consider each core as a unit in my work. If I want to do
> so for my chip (the one with 4 layers and 16 cores in each one), how does
> it affect on grid size? In order to generate a single temperature for each
> core, should I define the grid size as 4*4? Unfortunately the concept of
> grid is still a little vague to me.
>

>>>  HotSpot converts grid temperatures to floorplan block temperatures for
you. You may want to try different grid sizes (for example, 4x4, 8x8,
16x16) and find the grid size beyond which no significant accuracy is
achieved.

>
>
> 4. In the lcf file, is the layer with number 0, the nearest one to the
> heat sink or the farthest one?
>

>>> I think layer 0 is the farthest from the heatsink. You may want to
confirm that by looking at the code.

-Wei
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120304/f7495c3c/attachment.html 

From nedahpg at gmail.com  Tue Mar  6 23:06:18 2012
From: nedahpg at gmail.com (Neda Hassanpour)
Date: Wed, 7 Mar 2012 10:36:18 +0330
Subject: [Hotspot] (no subject)
In-Reply-To: <CANYeXBpm3Ca=jUCuUtXAyBu50S6_Sz-6Xc7Bynx1_WrKEFXKWg@mail.gmail.com>
References: <CANYeXBqa8RkUYuSqQ5GADCG8R1JVR-_jK6ykgiKpdozm4Wbi7Q@mail.gmail.com>
	<CAPa2X=TQ6VsU+-kWfYffOCfbfp-JTOWoPi=iNuiS1tn-b3yRuw@mail.gmail.com>
	<CANYeXBpm3Ca=jUCuUtXAyBu50S6_Sz-6Xc7Bynx1_WrKEFXKWg@mail.gmail.com>
Message-ID: <CAPa2X=Ss69RpYzgV6G6_NYhUGUVk2U5Aj_0E7eEuO2vzoNSUHQ@mail.gmail.com>

Hi,


Thanks for your help. Despite your helpful responses I have other questions
related to 3D stacking and I will be really grateful if you help me again.


1. My first question is about the concept of initial temperature (the
content of example.ttrace file) in 3D stacking simulation.

In my case which I?m going to consider each core in a layer as a unit, does
the initial temperature of each core suppose to demonstrate its temperature
before the execution of a specific task --with defined power consumption in
example.ptrace file-- on it? If I am wrong please let me know the right
answer.



2.To initialize the example.ttrace file in 3D stacking,  I should run a
steady-state simulation first to set the steady-state temperature as
initial temperature. Despite your response to my second question in my last
email, I think I cannot use commands below to run a steady-state simulation
for 3D stacking since these commands are for 2D floorplans.

*hotspot -c hotspot.config -f ev6.flp -p gcc.ptrace \*

*      -steady_file gcc.steady -model_type grid \ *

*       -grid_steady_file gcc.grid.steady*

* *

How can I run a steady-state simulation in 3D floorplans to set the initial
temperatures?




3.If I set the grid size 4*4, does it mean that HotSpot will  consider 16
grid cells in *each* layer?




4.If my assumption in question 3 is true, with 16 cores in each layer and
setting grid size as 4*4, Is it true to say that each grid cell temperature
which results after 3D stacking simulation, corresponds  to a core
temperature?




5.Using the commands below, how can I output the simulation results
(resulted temperatures) onto a specific file?

*hotspot -c hotspot.config -f <some_random_file> -p example.ptrace \*

*           -o example.ttrace -model_type grid \*

*           -grid_layer_file example.lcf*



Thank you so much

Neda







On Sun, Mar 4, 2012 at 20:13, Wei Huang <wh6p at virginia.edu> wrote:

>
>> 1. I understood that in gcc.grid.steady file, each line corresponds to a
>> grid cell and its steady-state temperature. So, is it true that the number
>> of lines in this file is determined by the size of grid which is defined in
>> the hotspot.config file before? (in this example this size is 64*64)
>>
>
> >>> yes.
>
>>
>>
>> 2. According to your response to my fifth question, I should run a
>> steady-state simulation first and set the steady-state temperature as my
>> initial temperature. So, what I think I should do now, Is running the
>> commands below, generating the gcc.grid.steady file, and using it for 3D
>> simulation as a list of initial temperature, is this true?
>>
>>
>>
>>        hotspot -c hotspot.config -f ev6.flp -p gcc.ptrace \
>>
>>       -steady_file gcc.steady -model_type grid \
>>
>        -grid_steady_file gcc.grid.steady
>>
>>
>>
>  >>> Looks good to me.
>
>  3. In my work I don?t need sub-core architectural blocks thermal details
>> and the only thing that is important to me is the cores temperature. So, I
>> think I should consider each core as a unit in my work. If I want to do
>> so for my chip (the one with 4 layers and 16 cores in each one), how does
>> it affect on grid size? In order to generate a single temperature for each
>> core, should I define the grid size as 4*4? Unfortunately the concept of
>> grid is still a little vague to me.
>>
>
> >>>  HotSpot converts grid temperatures to floorplan block temperatures
> for you. You may want to try different grid sizes (for example, 4x4, 8x8,
> 16x16) and find the grid size beyond which no significant accuracy is
> achieved.
>
>>
>>
>> 4. In the lcf file, is the layer with number 0, the nearest one to the
>> heat sink or the farthest one?
>>
>
> >>> I think layer 0 is the farthest from the heatsink. You may want to
> confirm that by looking at the code.
>
> -Wei
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120307/845186a5/attachment-0001.html 

From nedahpg at gmail.com  Wed Mar  7 10:33:30 2012
From: nedahpg at gmail.com (Neda Hassanpour)
Date: Wed, 7 Mar 2012 22:03:30 +0330
Subject: [Hotspot] Simulation problem
Message-ID: <CAPa2X=R=hyECP7BX=qeOgLhXJPno13afq2ivaReYS-L8q44bbQ@mail.gmail.com>

Hi,


I have a question about 3D stacking simulation. I have defined my own flp
and lcf files . after executing commands below  with proper lcf and flp
files, HotSpot returns ?-nan? instead of each unit temperature.

*hotspot -c hotspot.config -f <some_random_file> -p example.ptrace \*

*           -o example.ttrace -model_type grid \*

*           -grid_layer_file example.lcf*



Can you please help me to understand what causes this problem? What does a
word *nan* stand for?


Best regards
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120307/4b12ea81/attachment.html 

From wh6p at virginia.edu  Wed Mar  7 11:15:03 2012
From: wh6p at virginia.edu (Wei Huang)
Date: Wed, 7 Mar 2012 13:15:03 -0600
Subject: [Hotspot] Simulation problem
In-Reply-To: <CAPa2X=R=hyECP7BX=qeOgLhXJPno13afq2ivaReYS-L8q44bbQ@mail.gmail.com>
References: <CAPa2X=R=hyECP7BX=qeOgLhXJPno13afq2ivaReYS-L8q44bbQ@mail.gmail.com>
Message-ID: <CANYeXBqGyBiTP=4EBDkR+uigdjXND=UrMrizxA4kw5CUJmx46A@mail.gmail.com>

some usual causes of this problem could be:
- undefined (blank) area in the floorplan file, or overlapping block in the
floorplan
- chip size should be smaller than heat spreader size, which should be
smaller than heat sink size.

Wei

On Wed, Mar 7, 2012 at 12:33 PM, Neda Hassanpour <nedahpg at gmail.com> wrote:

> Hi,
>
>
> I have a question about 3D stacking simulation. I have defined my own flp
> and lcf files . after executing commands below  with proper lcf and flp
> files, HotSpot returns ?-nan? instead of each unit temperature.
>
> *hotspot -c hotspot.config -f <some_random_file> -p example.ptrace \*
>
> *           -o example.ttrace -model_type grid \*
>
> *           -grid_layer_file example.lcf*
>
>
>
> Can you please help me to understand what causes this problem? What does a
> word *nan* stand for?
>
>
> Best regards
>
>
> _______________________________________________
> 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/20120307/c798929e/attachment.html 

From nedahpg at gmail.com  Wed Mar  7 23:00:04 2012
From: nedahpg at gmail.com (Neda Hassanpour)
Date: Thu, 8 Mar 2012 10:30:04 +0330
Subject: [Hotspot] Simulation problem
In-Reply-To: <CANYeXBqGyBiTP=4EBDkR+uigdjXND=UrMrizxA4kw5CUJmx46A@mail.gmail.com>
References: <CAPa2X=R=hyECP7BX=qeOgLhXJPno13afq2ivaReYS-L8q44bbQ@mail.gmail.com>
	<CANYeXBqGyBiTP=4EBDkR+uigdjXND=UrMrizxA4kw5CUJmx46A@mail.gmail.com>
Message-ID: <CAPa2X=SrAXzSiDX7QGxZZrZuVYqkEVEdgsZHwwARwzxxVP=VsQ@mail.gmail.com>

*Hi,

Thanks for your help,
I checked my floorplan file. there was overlapping between blocks in the
floorplan.
but after editing  the floorplan file I still have the  same
problem(HotSpot returns -nan). I have attached  a sample of my flp file to
this email (and also others like lcf file and ...). can you please take a
look at them and let me know if there is a problem in defining these files?
can be any other reason for such a problem?

Thanks again
*
On Wed, Mar 7, 2012 at 22:45, Wei Huang <wh6p at virginia.edu> wrote:

> some usual causes of this problem could be:
> - undefined (blank) area in the floorplan file, or overlapping block in
> the floorplan
> - chip size should be smaller than heat spreader size, which should be
> smaller than heat sink size.
>
> Wei
>
> On Wed, Mar 7, 2012 at 12:33 PM, Neda Hassanpour <nedahpg at gmail.com>wrote:
>
>> Hi,
>>
>>
>> I have a question about 3D stacking simulation. I have defined my own flp
>> and lcf files . after executing commands below  with proper lcf and flp
>> files, HotSpot returns ?-nan? instead of each unit temperature.
>>
>> *hotspot -c hotspot.config -f <some_random_file> -p example.ptrace \*
>>
>> *           -o example.ttrace -model_type grid \*
>>
>> *           -grid_layer_file example.lcf*
>>
>>
>>
>> Can you please help me to understand what causes this problem? What does
>> a word *nan* stand for?
>>
>>
>> Best regards
>>
>>
>> _______________________________________________
>> 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/20120308/21fc1b48/attachment.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Files.tar.gz
Type: application/x-gzip
Size: 1200 bytes
Desc: not available
Url : http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120308/21fc1b48/attachment.bin 

From davizhao at gmail.com  Thu Mar  8 00:03:32 2012
From: davizhao at gmail.com (Dali Zhao)
Date: Thu, 8 Mar 2012 00:03:32 -0800
Subject: [Hotspot] Simulation problem
In-Reply-To: <CAPa2X=SrAXzSiDX7QGxZZrZuVYqkEVEdgsZHwwARwzxxVP=VsQ@mail.gmail.com>
References: <CAPa2X=R=hyECP7BX=qeOgLhXJPno13afq2ivaReYS-L8q44bbQ@mail.gmail.com>
	<CANYeXBqGyBiTP=4EBDkR+uigdjXND=UrMrizxA4kw5CUJmx46A@mail.gmail.com>
	<CAPa2X=SrAXzSiDX7QGxZZrZuVYqkEVEdgsZHwwARwzxxVP=VsQ@mail.gmail.com>
Message-ID: <CAE9BO94+EuMV+6qv8Jma+P0rR=18zR4WRnK+k9aMYJw0t6jXPA@mail.gmail.com>

Hi,
  Did you check the config file? You need to make sure the chip size is
smaller than the heat spreader size, whish should be smaller than the heat
sink size.

David

On Wed, Mar 7, 2012 at 11:00 PM, Neda Hassanpour <nedahpg at gmail.com> wrote:

> *Hi,
>
> Thanks for your help,
> I checked my floorplan file. there was overlapping between blocks in the
> floorplan.
> but after editing  the floorplan file I still have the  same
> problem(HotSpot returns -nan). I have attached  a sample of my flp file to
> this email (and also others like lcf file and ...). can you please take a
> look at them and let me know if there is a problem in defining these files?
> can be any other reason for such a problem?
>
> Thanks again
> *
>
> On Wed, Mar 7, 2012 at 22:45, Wei Huang <wh6p at virginia.edu> wrote:
>
>> some usual causes of this problem could be:
>> - undefined (blank) area in the floorplan file, or overlapping block in
>> the floorplan
>> - chip size should be smaller than heat spreader size, which should be
>> smaller than heat sink size.
>>
>> Wei
>>
>> On Wed, Mar 7, 2012 at 12:33 PM, Neda Hassanpour <nedahpg at gmail.com>wrote:
>>
>>> Hi,
>>>
>>>
>>> I have a question about 3D stacking simulation. I have defined my own
>>> flp and lcf files . after executing commands below  with proper lcf and flp
>>> files, HotSpot returns ?-nan? instead of each unit temperature.
>>>
>>> *hotspot -c hotspot.config -f <some_random_file> -p example.ptrace \*
>>>
>>> *           -o example.ttrace -model_type grid \*
>>>
>>> *           -grid_layer_file example.lcf*
>>>
>>>
>>>
>>> Can you please help me to understand what causes this problem? What does
>>> a word *nan* stand for?
>>>
>>>
>>> Best regards
>>>
>>>
>>> _______________________________________________
>>> 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
>
>


-- 
Dali Zhao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120308/a88d83ca/attachment.html 

From nedahpg at gmail.com  Thu Mar  8 00:19:14 2012
From: nedahpg at gmail.com (Neda Hassanpour)
Date: Thu, 8 Mar 2012 11:49:14 +0330
Subject: [Hotspot] Simulation problem
In-Reply-To: <CAE9BO94+EuMV+6qv8Jma+P0rR=18zR4WRnK+k9aMYJw0t6jXPA@mail.gmail.com>
References: <CAPa2X=R=hyECP7BX=qeOgLhXJPno13afq2ivaReYS-L8q44bbQ@mail.gmail.com>
	<CANYeXBqGyBiTP=4EBDkR+uigdjXND=UrMrizxA4kw5CUJmx46A@mail.gmail.com>
	<CAPa2X=SrAXzSiDX7QGxZZrZuVYqkEVEdgsZHwwARwzxxVP=VsQ@mail.gmail.com>
	<CAE9BO94+EuMV+6qv8Jma+P0rR=18zR4WRnK+k9aMYJw0t6jXPA@mail.gmail.com>
Message-ID: <CAPa2X=SkOvsuwZEa92D+V1K4Ogo__sCwHAQyYmCGW9N6FOedcw@mail.gmail.com>

*Hi,*
*
*
*when you say chip size, do you mean the parameter chip thickness in the
config file?*
*If you mean so, yes,  I have checked them, their values are as below.*
*
*
*chip thickness=0.000011929*
*heatsink thickness=0.0069*
*spreader thickness=0.001*
*
*
*when I downloaded HotSpot, I did not change the values in the config file
except the grid size and **chip thickness.*
*Is it necessary to reset any other parameter in the config file  if want
to use it for my own flp and lcf files in grid model? *

On Thu, Mar 8, 2012 at 11:33, Dali Zhao <davizhao at gmail.com> wrote:

> Hi,
>   Did you check the config file? You need to make sure the chip size is
> smaller than the heat spreader size, whish should be smaller than the heat
> sink size.
>
> David
>
>
> On Wed, Mar 7, 2012 at 11:00 PM, Neda Hassanpour <nedahpg at gmail.com>wrote:
>
>> *Hi,
>>
>> Thanks for your help,
>> I checked my floorplan file. there was overlapping between blocks in the
>> floorplan.
>> but after editing  the floorplan file I still have the  same
>> problem(HotSpot returns -nan). I have attached  a sample of my flp file to
>> this email (and also others like lcf file and ...). can you please take a
>> look at them and let me know if there is a problem in defining these files?
>> can be any other reason for such a problem?
>>
>> Thanks again
>> *
>>
>> On Wed, Mar 7, 2012 at 22:45, Wei Huang <wh6p at virginia.edu> wrote:
>>
>>> some usual causes of this problem could be:
>>> - undefined (blank) area in the floorplan file, or overlapping block in
>>> the floorplan
>>> - chip size should be smaller than heat spreader size, which should be
>>> smaller than heat sink size.
>>>
>>> Wei
>>>
>>> On Wed, Mar 7, 2012 at 12:33 PM, Neda Hassanpour <nedahpg at gmail.com>wrote:
>>>
>>>> Hi,
>>>>
>>>>
>>>> I have a question about 3D stacking simulation. I have defined my own
>>>> flp and lcf files . after executing commands below  with proper lcf and flp
>>>> files, HotSpot returns ?-nan? instead of each unit temperature.
>>>>
>>>> *hotspot -c hotspot.config -f <some_random_file> -p example.ptrace \*
>>>>
>>>> *           -o example.ttrace -model_type grid \*
>>>>
>>>> *           -grid_layer_file example.lcf*
>>>>
>>>>
>>>>
>>>> Can you please help me to understand what causes this problem? What
>>>> does a word *nan* stand for?
>>>>
>>>>
>>>> Best regards
>>>>
>>>>
>>>> _______________________________________________
>>>> 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
>>
>>
>
>
> --
> Dali Zhao
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120308/d1ace78f/attachment-0001.html 

From davizhao at gmail.com  Thu Mar  8 00:29:44 2012
From: davizhao at gmail.com (Dali Zhao)
Date: Thu, 8 Mar 2012 00:29:44 -0800
Subject: [Hotspot] Simulation problem
In-Reply-To: <CAPa2X=SkOvsuwZEa92D+V1K4Ogo__sCwHAQyYmCGW9N6FOedcw@mail.gmail.com>
References: <CAPa2X=R=hyECP7BX=qeOgLhXJPno13afq2ivaReYS-L8q44bbQ@mail.gmail.com>
	<CANYeXBqGyBiTP=4EBDkR+uigdjXND=UrMrizxA4kw5CUJmx46A@mail.gmail.com>
	<CAPa2X=SrAXzSiDX7QGxZZrZuVYqkEVEdgsZHwwARwzxxVP=VsQ@mail.gmail.com>
	<CAE9BO94+EuMV+6qv8Jma+P0rR=18zR4WRnK+k9aMYJw0t6jXPA@mail.gmail.com>
	<CAPa2X=SkOvsuwZEa92D+V1K4Ogo__sCwHAQyYmCGW9N6FOedcw@mail.gmail.com>
Message-ID: <CAE9BO94Mv-4t_b-oB8fv0PXdjQOPzAM87LSBSQKVx5fRWX5WBg@mail.gmail.com>

The thickness is not the problem. Check the heat sink side (s_sink) and
spreader side (s_spreader).
When you use a bigger die area in the flp file, you need to adjust heat
sink side and spreader side.

On Thu, Mar 8, 2012 at 12:19 AM, Neda Hassanpour <nedahpg at gmail.com> wrote:

> *Hi,*
> *
> *
> *when you say chip size, do you mean the parameter chip thickness in the
> config file?*
> *If you mean so, yes,  I have checked them, their values are as below.*
> *
> *
> *chip thickness=0.000011929*
> *heatsink thickness=0.0069*
> *spreader thickness=0.001*
> *
> *
> *when I downloaded HotSpot, I did not change the values in the config
> file except the grid size and **chip thickness.*
> *Is it necessary to reset any other parameter in the config file  if want
> to use it for my own flp and lcf files in grid model? *
>
> On Thu, Mar 8, 2012 at 11:33, Dali Zhao <davizhao at gmail.com> wrote:
>
>> Hi,
>>   Did you check the config file? You need to make sure the chip size is
>> smaller than the heat spreader size, whish should be smaller than the heat
>> sink size.
>>
>> David
>>
>>
>> On Wed, Mar 7, 2012 at 11:00 PM, Neda Hassanpour <nedahpg at gmail.com>wrote:
>>
>>> *Hi,
>>>
>>> Thanks for your help,
>>> I checked my floorplan file. there was overlapping between blocks in the
>>> floorplan.
>>> but after editing  the floorplan file I still have the  same
>>> problem(HotSpot returns -nan). I have attached  a sample of my flp file to
>>> this email (and also others like lcf file and ...). can you please take a
>>> look at them and let me know if there is a problem in defining these files?
>>> can be any other reason for such a problem?
>>>
>>> Thanks again
>>> *
>>>
>>> On Wed, Mar 7, 2012 at 22:45, Wei Huang <wh6p at virginia.edu> wrote:
>>>
>>>> some usual causes of this problem could be:
>>>> - undefined (blank) area in the floorplan file, or overlapping block in
>>>> the floorplan
>>>> - chip size should be smaller than heat spreader size, which should be
>>>> smaller than heat sink size.
>>>>
>>>> Wei
>>>>
>>>> On Wed, Mar 7, 2012 at 12:33 PM, Neda Hassanpour <nedahpg at gmail.com>wrote:
>>>>
>>>>> Hi,
>>>>>
>>>>>
>>>>> I have a question about 3D stacking simulation. I have defined my own
>>>>> flp and lcf files . after executing commands below  with proper lcf and flp
>>>>> files, HotSpot returns ?-nan? instead of each unit temperature.
>>>>>
>>>>> *hotspot -c hotspot.config -f <some_random_file> -p example.ptrace \*
>>>>>
>>>>> *           -o example.ttrace -model_type grid \*
>>>>>
>>>>> *           -grid_layer_file example.lcf*
>>>>>
>>>>>
>>>>>
>>>>> Can you please help me to understand what causes this problem? What
>>>>> does a word *nan* stand for?
>>>>>
>>>>>
>>>>> Best regards
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> 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
>>>
>>>
>>
>>
>> --
>> Dali Zhao
>>
>>
>


-- 
Dali Zhao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120308/b1e713df/attachment.html 

From nedahpg at gmail.com  Thu Mar  8 00:45:42 2012
From: nedahpg at gmail.com (Neda Hassanpour)
Date: Thu, 8 Mar 2012 12:15:42 +0330
Subject: [Hotspot] Simulation problem
In-Reply-To: <CAE9BO94Mv-4t_b-oB8fv0PXdjQOPzAM87LSBSQKVx5fRWX5WBg@mail.gmail.com>
References: <CAPa2X=R=hyECP7BX=qeOgLhXJPno13afq2ivaReYS-L8q44bbQ@mail.gmail.com>
	<CANYeXBqGyBiTP=4EBDkR+uigdjXND=UrMrizxA4kw5CUJmx46A@mail.gmail.com>
	<CAPa2X=SrAXzSiDX7QGxZZrZuVYqkEVEdgsZHwwARwzxxVP=VsQ@mail.gmail.com>
	<CAE9BO94+EuMV+6qv8Jma+P0rR=18zR4WRnK+k9aMYJw0t6jXPA@mail.gmail.com>
	<CAPa2X=SkOvsuwZEa92D+V1K4Ogo__sCwHAQyYmCGW9N6FOedcw@mail.gmail.com>
	<CAE9BO94Mv-4t_b-oB8fv0PXdjQOPzAM87LSBSQKVx5fRWX5WBg@mail.gmail.com>
Message-ID: <CAPa2X=QfpUNWe_xa5d+h4B6HDp9cHX+ZK637t5XXx+3AEFcTOw@mail.gmail.com>

So if I have got it in the right way, I should adjust s_spreader and s_sink
parameters whose default values in
the config file is now 0.03 and 0.06 respectively.
which parameter represents the chip size?
Is there any specific rule  to change these values(s_spreader, s_sink)
according to my floorplan?
Is there any other parameters in the config file which I have to change
according to my flp file?

On Thu, Mar 8, 2012 at 11:59, Dali Zhao <davizhao at gmail.com> wrote:

> The thickness is not the problem. Check the heat sink side (s_sink) and
> spreader side (s_spreader).
> When you use a bigger die area in the flp file, you need to adjust heat
> sink side and spreader side.
>
>
> On Thu, Mar 8, 2012 at 12:19 AM, Neda Hassanpour <nedahpg at gmail.com>wrote:
>
>> *Hi,*
>> *
>> *
>> *when you say chip size, do you mean the parameter chip thickness in the
>> config file?*
>> *If you mean so, yes,  I have checked them, their values are as below.*
>>  *
>> *
>> *chip thickness=0.000011929*
>> *heatsink thickness=0.0069*
>> *spreader thickness=0.001*
>> *
>> *
>> *when I downloaded HotSpot, I did not change the values in the config
>> file except the grid size and **chip thickness.*
>> *Is it necessary to reset any other parameter in the config file  if
>> want to use it for my own flp and lcf files in grid model? *
>>
>> On Thu, Mar 8, 2012 at 11:33, Dali Zhao <davizhao at gmail.com> wrote:
>>
>>> Hi,
>>>   Did you check the config file? You need to make sure the chip size is
>>> smaller than the heat spreader size, whish should be smaller than the heat
>>> sink size.
>>>
>>> David
>>>
>>>
>>> On Wed, Mar 7, 2012 at 11:00 PM, Neda Hassanpour <nedahpg at gmail.com>wrote:
>>>
>>>> *Hi,
>>>>
>>>> Thanks for your help,
>>>> I checked my floorplan file. there was overlapping between blocks in
>>>> the floorplan.
>>>> but after editing  the floorplan file I still have the  same
>>>> problem(HotSpot returns -nan). I have attached  a sample of my flp file to
>>>> this email (and also others like lcf file and ...). can you please take a
>>>> look at them and let me know if there is a problem in defining these files?
>>>> can be any other reason for such a problem?
>>>>
>>>> Thanks again
>>>> *
>>>>
>>>> On Wed, Mar 7, 2012 at 22:45, Wei Huang <wh6p at virginia.edu> wrote:
>>>>
>>>>> some usual causes of this problem could be:
>>>>> - undefined (blank) area in the floorplan file, or overlapping block
>>>>> in the floorplan
>>>>> - chip size should be smaller than heat spreader size, which should be
>>>>> smaller than heat sink size.
>>>>>
>>>>> Wei
>>>>>
>>>>> On Wed, Mar 7, 2012 at 12:33 PM, Neda Hassanpour <nedahpg at gmail.com>wrote:
>>>>>
>>>>>> Hi,
>>>>>>
>>>>>>
>>>>>> I have a question about 3D stacking simulation. I have defined my own
>>>>>> flp and lcf files . after executing commands below  with proper lcf and flp
>>>>>> files, HotSpot returns ?-nan? instead of each unit temperature.
>>>>>>
>>>>>> *hotspot -c hotspot.config -f <some_random_file> -p example.ptrace \*
>>>>>>
>>>>>> *           -o example.ttrace -model_type grid \*
>>>>>>
>>>>>> *           -grid_layer_file example.lcf*
>>>>>>
>>>>>>
>>>>>>
>>>>>> Can you please help me to understand what causes this problem? What
>>>>>> does a word *nan* stand for?
>>>>>>
>>>>>>
>>>>>> Best regards
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> 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
>>>>
>>>>
>>>
>>>
>>> --
>>> Dali Zhao
>>>
>>>
>>
>
>
> --
> Dali Zhao
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120308/d7eccc24/attachment.html 

From kh15 at rice.edu  Thu Mar  8 11:09:07 2012
From: kh15 at rice.edu (Kangqiao Hu)
Date: Thu, 8 Mar 2012 13:09:07 -0600
Subject: [Hotspot] non-architectural power trace
Message-ID: <CAHC2y90mkjwLw20+WMLOp1maXbdMz4YTuKg_75DC3DXJCnLfBQ@mail.gmail.com>

Hi,

I am recently learning and using hotspot for thermal simulation. I've a
question about power trace. I know that Wattch/SimpleSim or other
architectural power simulators can generate appropriate power
trace information for hotspot. However, I am using Hotspot for
non-architectural simulations, for instance JPEG encoder chip. I have
access to Synopsys tools and Cadence. I want to know if these tools can
also give the power information that can be fed to Hotspot? If so, which
application of Synopsys/Cadence should I use? Thanks very much.

Best regards.

-- 
Kangqiao Hu
Dept. of Electrical and Computer Engineering
Rice University
Tel: 8322719006
Email: kangqiaohu at gmail.com
Alternate E-mail: kh15 at rice.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120308/f6b09c55/attachment.html 

From nedahpg at gmail.com  Sat Mar 10 08:54:53 2012
From: nedahpg at gmail.com (Neda Hassanpour)
Date: Sat, 10 Mar 2012 20:24:53 +0330
Subject: [Hotspot] Simulation problem
In-Reply-To: <CAPa2X=QfpUNWe_xa5d+h4B6HDp9cHX+ZK637t5XXx+3AEFcTOw@mail.gmail.com>
References: <CAPa2X=R=hyECP7BX=qeOgLhXJPno13afq2ivaReYS-L8q44bbQ@mail.gmail.com>
	<CANYeXBqGyBiTP=4EBDkR+uigdjXND=UrMrizxA4kw5CUJmx46A@mail.gmail.com>
	<CAPa2X=SrAXzSiDX7QGxZZrZuVYqkEVEdgsZHwwARwzxxVP=VsQ@mail.gmail.com>
	<CAE9BO94+EuMV+6qv8Jma+P0rR=18zR4WRnK+k9aMYJw0t6jXPA@mail.gmail.com>
	<CAPa2X=SkOvsuwZEa92D+V1K4Ogo__sCwHAQyYmCGW9N6FOedcw@mail.gmail.com>
	<CAE9BO94Mv-4t_b-oB8fv0PXdjQOPzAM87LSBSQKVx5fRWX5WBg@mail.gmail.com>
	<CAPa2X=QfpUNWe_xa5d+h4B6HDp9cHX+ZK637t5XXx+3AEFcTOw@mail.gmail.com>
Message-ID: <CAPa2X=RY47kPuEfZSBwH4kbF9ByrOwbJbOh--0KAqCJj2qZVYA@mail.gmail.com>

*
*
---------- Forwarded message ----------
From: Neda Hassanpour <nedahpg at gmail.com>
Date: Thu, Mar 8, 2012 at 12:15
Subject: Re: [Hotspot] Simulation problem
To: Dali Zhao <davizhao at gmail.com>
Cc: Wei Huang <wh6p at virginia.edu>, hotspot <hotspot at cs.virginia.edu>



*So if I have got it in the right way, I should adjust s_spreader and
s_sink parameters whose default values in *
*the config file is now 0.03 and 0.06 respectively.*
*which parameter represents the chip size?*
*Is there any specific rule  to change these values(s_spreader, s_sink)
according to my floorplan?*
*Is there any other parameters in the config file which I have to change
according to my flp file?*


On Thu, Mar 8, 2012 at 11:59, Dali Zhao <davizhao at gmail.com> wrote:

> The thickness is not the problem. Check the heat sink side (s_sink) and
> spreader side (s_spreader).
> When you use a bigger die area in the flp file, you need to adjust heat
> sink side and spreader side.
>
>
> On Thu, Mar 8, 2012 at 12:19 AM, Neda Hassanpour <nedahpg at gmail.com>wrote:
>
>> *Hi,*
>> *
>> *
>> *when you say chip size, do you mean the parameter chip thickness in the
>> config file?*
>> *If you mean so, yes,  I have checked them, their values are as below.*
>>  *
>> *
>> *chip thickness=0.000011929*
>> *heatsink thickness=0.0069*
>> *spreader thickness=0.001*
>> *
>> *
>> *when I downloaded HotSpot, I did not change the values in the config
>> file except the grid size and **chip thickness.*
>> *Is it necessary to reset any other parameter in the config file  if
>> want to use it for my own flp and lcf files in grid model? *
>>
>> On Thu, Mar 8, 2012 at 11:33, Dali Zhao <davizhao at gmail.com> wrote:
>>
>>> Hi,
>>>   Did you check the config file? You need to make sure the chip size is
>>> smaller than the heat spreader size, whish should be smaller than the heat
>>> sink size.
>>>
>>> David
>>>
>>>
>>> On Wed, Mar 7, 2012 at 11:00 PM, Neda Hassanpour <nedahpg at gmail.com>wrote:
>>>
>>>> *Hi,
>>>>
>>>> Thanks for your help,
>>>> I checked my floorplan file. there was overlapping between blocks in
>>>> the floorplan.
>>>> but after editing  the floorplan file I still have the  same
>>>> problem(HotSpot returns -nan). I have attached  a sample of my flp file to
>>>> this email (and also others like lcf file and ...). can you please take a
>>>> look at them and let me know if there is a problem in defining these files?
>>>> can be any other reason for such a problem?
>>>>
>>>> Thanks again
>>>> *
>>>>
>>>> On Wed, Mar 7, 2012 at 22:45, Wei Huang <wh6p at virginia.edu> wrote:
>>>>
>>>>> some usual causes of this problem could be:
>>>>> - undefined (blank) area in the floorplan file, or overlapping block
>>>>> in the floorplan
>>>>> - chip size should be smaller than heat spreader size, which should be
>>>>> smaller than heat sink size.
>>>>>
>>>>> Wei
>>>>>
>>>>> On Wed, Mar 7, 2012 at 12:33 PM, Neda Hassanpour <nedahpg at gmail.com>wrote:
>>>>>
>>>>>> Hi,
>>>>>>
>>>>>>
>>>>>> I have a question about 3D stacking simulation. I have defined my own
>>>>>> flp and lcf files . after executing commands below  with proper lcf and flp
>>>>>> files, HotSpot returns ?-nan? instead of each unit temperature.
>>>>>>
>>>>>> *hotspot -c hotspot.config -f <some_random_file> -p example.ptrace \*
>>>>>>
>>>>>> *           -o example.ttrace -model_type grid \*
>>>>>>
>>>>>> *           -grid_layer_file example.lcf*
>>>>>>
>>>>>>
>>>>>>
>>>>>> Can you please help me to understand what causes this problem? What
>>>>>> does a word *nan* stand for?
>>>>>>
>>>>>>
>>>>>> Best regards
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> 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
>>>>
>>>>
>>>
>>>
>>> --
>>> Dali Zhao
>>>
>>>
>>
>
>
> --
> Dali Zhao
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120310/649ed83f/attachment-0001.html 

From wh6p at virginia.edu  Sat Mar 10 09:54:11 2012
From: wh6p at virginia.edu (Wei Huang)
Date: Sat, 10 Mar 2012 11:54:11 -0600
Subject: [Hotspot] Simulation problem
In-Reply-To: <CAPa2X=RY47kPuEfZSBwH4kbF9ByrOwbJbOh--0KAqCJj2qZVYA@mail.gmail.com>
References: <CAPa2X=R=hyECP7BX=qeOgLhXJPno13afq2ivaReYS-L8q44bbQ@mail.gmail.com>
	<CANYeXBqGyBiTP=4EBDkR+uigdjXND=UrMrizxA4kw5CUJmx46A@mail.gmail.com>
	<CAPa2X=SrAXzSiDX7QGxZZrZuVYqkEVEdgsZHwwARwzxxVP=VsQ@mail.gmail.com>
	<CAE9BO94+EuMV+6qv8Jma+P0rR=18zR4WRnK+k9aMYJw0t6jXPA@mail.gmail.com>
	<CAPa2X=SkOvsuwZEa92D+V1K4Ogo__sCwHAQyYmCGW9N6FOedcw@mail.gmail.com>
	<CAE9BO94Mv-4t_b-oB8fv0PXdjQOPzAM87LSBSQKVx5fRWX5WBg@mail.gmail.com>
	<CAPa2X=QfpUNWe_xa5d+h4B6HDp9cHX+ZK637t5XXx+3AEFcTOw@mail.gmail.com>
	<CAPa2X=RY47kPuEfZSBwH4kbF9ByrOwbJbOh--0KAqCJj2qZVYA@mail.gmail.com>
Message-ID: <CANYeXBrzLavjjuRMYQQnpLCsMrj1Uxo_5U-7nzRXq3j2p+GVvA@mail.gmail.com>

On Sat, Mar 10, 2012 at 10:54 AM, Neda Hassanpour <nedahpg at gmail.com> wrote:

> *
> *
> ---------- Forwarded message ----------
> From: Neda Hassanpour <nedahpg at gmail.com>
> Date: Thu, Mar 8, 2012 at 12:15
> Subject: Re: [Hotspot] Simulation problem
> To: Dali Zhao <davizhao at gmail.com>
> Cc: Wei Huang <wh6p at virginia.edu>, hotspot <hotspot at cs.virginia.edu>
>
>
>
> *So if I have got it in the right way, I should adjust s_spreader and
> s_sink parameters whose default values in *
> *the config file is now 0.03 and 0.06 respectively.*
> *which parameter represents the chip size?*
> *Is there any specific rule  to change these values(s_spreader, s_sink)
> according to my floorplan?*
> *Is there any other parameters in the config file which I have to change
> according to my flp file?*
>

The chip size is decided by your floorplan file, so you should make sure
that spreader and heatsink sizes are bigger than your chip size. The
default sizes for spreader and sink are 0.03m*0.03m, and 0.06m*0.06m,
respectively.

As to how to choose the "right" sizes, it really depends on what thermal
package you want to put your chip into.

Hope this helps.

-Wei
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120310/3e9332e0/attachment.html 

From arihant at iith.ac.in  Sat Mar 10 10:55:54 2012
From: arihant at iith.ac.in (Arihant Jain)
Date: Sat, 10 Mar 2012 10:55:54 -0800
Subject: [Hotspot] Modeling 3D IC with Hotspot
Message-ID: <CAB-vPZMN8KWSR8vDD9Qy7BakN4hjh8OfO4iuh9eDfa9dXRjVUw@mail.gmail.com>

Hello,

I am trying to generate the thermal profile for 2 stacked Alpha 21264 cores
using Grid model.
My Question is:

How can i generate the temperature corresponding to all the grid
structures(4096 for 64*64) for both the layers and Plot them?

Hotspot support that for a single layer at a time but in order to
incorporate the heat transfer from layers sitting above and below,we need
to calculate those grid temperature values simultaneously.

Regards
Arihant Jain
4th Year Student
Bachelor of Technology(Hons)
Department of Electrical Engineering
Indian Institute of Technology,Hyderabad
http://www.iith.ac.in

Ph.No:+919441919112
Email ID: arihant at iith.ac.in, arihantjain003 at gmail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120310/d3e4b6fa/attachment.html 

From skadron at cs.virginia.edu  Sun Mar 11 07:07:30 2012
From: skadron at cs.virginia.edu (Kevin Skadron)
Date: Sun, 11 Mar 2012 11:07:30 -0400
Subject: [Hotspot] non-architectural power trace
In-Reply-To: <CAHC2y90mkjwLw20+WMLOp1maXbdMz4YTuKg_75DC3DXJCnLfBQ@mail.gmail.com>
References: <CAHC2y90mkjwLw20+WMLOp1maXbdMz4YTuKg_75DC3DXJCnLfBQ@mail.gmail.com>
Message-ID: <4F5CBFB2.10306@cs.virginia.edu>

The only requirement is that the power trace include power for each 
block in the floorplan (or each grid cell).  HotSpot tries to be as 
portable as possible and hence as agnostic as possible about how the 
floorplan and power trace are obtained.
/K

On 3/8/2012 2:09 PM, Kangqiao Hu wrote:
> Hi,
> I am recently learning and using hotspot for thermal simulation. I've a
> question about power trace. I know that Wattch/SimpleSim or other
> architectural power simulators can generate appropriate power
> trace information for hotspot. However, I am using Hotspot for
> non-architectural simulations, for instance JPEG encoder chip. I have
> access to Synopsys tools and Cadence. I want to know if these tools can
> also give the power information that can be fed to Hotspot? If so, which
> application of Synopsys/Cadence should I use? Thanks very much.
> Best regards.
>
> --
> Kangqiao Hu
> Dept. of Electrical and Computer Engineering
> Rice University
> Tel: 8322719006
> Email: kangqiaohu at gmail.com <mailto:aohu at gmail.com>
> Alternate E-mail: kh15 at rice.edu <mailto:kh15 at rice.edu>
>
>
>
>
> _______________________________________________
> HotSpot mailing list
> HotSpot at mail.cs.virginia.edu
> http://www.cs.virginia.edu/mailman/listinfo/hotspot

From scorbetta at elet.polimi.it  Mon Mar 12 02:46:20 2012
From: scorbetta at elet.polimi.it (Simone Corbetta)
Date: Mon, 12 Mar 2012 11:46:20 +0100
Subject: [Hotspot] Modeling 3D IC with Hotspot
In-Reply-To: <CAB-vPZNpZSnJeBcy1w9rTa82U+LXMC+6+fsiDsqEoqCMNZfLYg@mail.gmail.com>
References: <CAB-vPZMN8KWSR8vDD9Qy7BakN4hjh8OfO4iuh9eDfa9dXRjVUw@mail.gmail.com>
	<CAH8TaY=vRSCxE2+2egc9bCW30worU2W-YeEy7uURwmrcsaQg=w@mail.gmail.com>
	<CAB-vPZNpZSnJeBcy1w9rTa82U+LXMC+6+fsiDsqEoqCMNZfLYg@mail.gmail.com>
Message-ID: <CAH8TaYn1gaUXFyC0jh3VCi8mm5KJofdx=pF7+JtZheqgewfsVA@mail.gmail.com>

Dear Arihant,

I believe HotSpot by default dumps out only the last (top) layer of
the 3D grid model. I think you can easily extend the code to dump out
the grid model in every layer. I remember once I came across the
definition of the 'cuboid' data structure in the 'grid_model_vector_t'
struct within 'temperature_grid.h' header. Maybe you can go and take a
look at it :-D

Let's stay in touch about this, I am interested in this, too..

Best reagards

On Sun, Mar 11, 2012 at 4:50 AM, Arihant Jain <arihant at iith.ac.in> wrote:
> Hey Simone...Thanks for your reply.
>
> I m already using the Layer Configuration file(.lcf) to describe various
> layers and thermal conductivity between them and it is also generating the
> temperature of all the blocks across all the ?layer.
>
> What I want is instead of associating one particular temperature to a
> block,it should give me the temperatures of all 64*64 grid so that i can
> plot it using the grid_thermal_map.pl.
>
> Temperature for the all the grid structures can be obtained by using the
> grid_steady_file for a single layer floor plan but the same command cannot
> be used for getting grid temperatures for Multiple layers considered in .lcf
> file.(I tried it)
>
> On Sat, Mar 10, 2012 at 3:26 PM, Simone Corbetta <scorbetta at elet.polimi.it>
> wrote:
>>
>> Hi Arihant,
>>
>> HotSpot has some support for 3D stacking. You should specify a layer
>> configuration file (LCF), as done in the example.lcf example file in
>> the HotSpot release. Major information can be found in
>> the "How To" section on the webpage. However, some tricks are required
>> to model vias.
>>
>> Best regards
>>
>> On Sat, Mar 10, 2012 at 7:55 PM, Arihant Jain <arihant at iith.ac.in> wrote:
>> > Hello,
>> >
>> > I am trying to generate the thermal profile for 2 stacked Alpha 21264
>> > cores
>> > using Grid model.
>> > My Question is:
>> >
>> > How can i generate the temperature corresponding to all the grid
>> > structures(4096 for 64*64) for both the layers and Plot them?
>> >
>> > Hotspot support that for a single layer at a time but in order to
>> > incorporate the heat transfer from layers sitting above and below,we
>> > need to
>> > calculate those grid temperature values?simultaneously.
>> >
>> > Regards
>> > Arihant Jain
>> > 4th Year Student
>> > Bachelor of Technology(Hons)
>> > Department of Electrical Engineering
>> > Indian Institute of Technology,Hyderabad
>> > http://www.iith.ac.in
>> >
>> > Ph.No:+919441919112
>> > Email ID: arihant at iith.ac.in, arihantjain003 at gmail.com
>> >
>> >
>> >
>> > _______________________________________________
>> > HotSpot mailing list
>> > HotSpot at mail.cs.virginia.edu
>> > http://www.cs.virginia.edu/mailman/listinfo/hotspot
>> >
>>
>>
>>
>> --
>> Simone Corbetta, Ph.D. student
>>
>> Politecnico di Milano
>> Dipartimento di Elettronica e Informazione
>> Viale delle Rimembranze di Lambrate 14
>> 20134 Milano, ITALY
>>
>> Office: +39.02.2399.9623
>> Home:?http://home.dei.polimi.it/scorbetta
>> Linkedin:?http://www.linkedin.com/in/thelibrarian
>
>
>
>
> --
>
> Arihant Jain
> 4th Year Student
> Bachelor of Technology(Hons)
> Department of Electrical Engineering
> Indian Institute of Technology,Hyderabad
> http://www.iith.ac.in
>
> Ph.No:+919441919112
> Email ID: arihant at iith.ac.in, arihantjain003 at gmail.com
>
>



-- 
Simone Corbetta, Ph.D. student

Politecnico di Milano
Dipartimento di Elettronica e Informazione
Viale delle Rimembranze di Lambrate 14
20134 Milano, ITALY

Office: +39.02.2399.9623
Home:?http://home.dei.polimi.it/scorbetta
Linkedin:?http://www.linkedin.com/in/thelibrarian


From solidrepellent at gmail.com  Tue Mar 13 02:18:06 2012
From: solidrepellent at gmail.com (Solid Repel)
Date: Tue, 13 Mar 2012 12:18:06 +0200
Subject: [Hotspot] Simplescalar benchmarks ?
Message-ID: <CAAYqiTu=9t02hssOp7c1MvbkyO7G7pNXHuLeQg781kd6nE8q4g@mail.gmail.com>

Hi all,

I am using PTscalar tool to generate the power trace. But I need proper
benchmarks in order to get something useful. Could you please let me know
the benchmarks that can be used with simplescalar which are free and/or
open-source.

Thanks and Regards,
Naveen.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120313/f49eb566/attachment.html 

From devendra.rai at tik.ee.ethz.ch  Tue Mar 13 03:05:52 2012
From: devendra.rai at tik.ee.ethz.ch (Devendra Rai)
Date: Tue, 13 Mar 2012 12:05:52 +0100
Subject: [Hotspot] Simplescalar benchmarks ?
In-Reply-To: <CAAYqiTu=9t02hssOp7c1MvbkyO7G7pNXHuLeQg781kd6nE8q4g@mail.gmail.com>
References: <CAAYqiTu=9t02hssOp7c1MvbkyO7G7pNXHuLeQg781kd6nE8q4g@mail.gmail.com>
Message-ID: <CAM3NbWEfkfbJcFgh1BoJQjd3w5tFYeHUjCzJLsg=XPGbtfa06g@mail.gmail.com>

Consider MiBench benchmarks here: http://www.eecs.umich.edu/mibench/

I was able to compile most of them for SimpleScalar/Wattch + Hotspot
toolchain.

Best
Devendra Rai


On Tue, Mar 13, 2012 at 11:18, Solid Repel <solidrepellent at gmail.com> wrote:

> Hi all,
>
> I am using PTscalar tool to generate the power trace. But I need proper
> benchmarks in order to get something useful. Could you please let me know
> the benchmarks that can be used with simplescalar which are free and/or
> open-source.
>
> Thanks and Regards,
> Naveen.
>
>
>
> _______________________________________________
> 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/20120313/8b273aeb/attachment.html 

From smillican at wisc.edu  Tue Mar 13 22:04:51 2012
From: smillican at wisc.edu (Spencer Millican)
Date: Wed, 14 Mar 2012 01:04:51 -0500
Subject: [Hotspot] Hotspot - Extracting Conductance Values
Message-ID: <CAFT6x5cOqx03VKek9voVxD5vc6ZiD0==iYuF4ruYVRWymOHZNA@mail.gmail.com>

Hotspot users,

I'm trying to use Hotspot to extract conductance values to use in
a separate program using a block model to calculate steady state
temperatures under different power conditions.  I'm certain the values I
want to extract calculated in temperature_block.c, and I just want to add a
print statement to give me the values I need and exit.

However, I'm having some troubles finding the values I want.  Can anyone
help point me to these values?

-Conductance/Resistance between two blocks in the silicon layer
-Conductance/Resistance between two blocks in the HS layer
-Conductance/Resistance between a block in the silicon layer and
its corresponding HS layer block
-Conductance/Resistance between a block in the HS layer and ambient

Can someone please point me out to the variables containing these values?
 It would be greatly appreciated.

-Spencer Millican
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120314/c80645d6/attachment-0001.html 

From raidalzobaidi at yahoo.com  Mon Mar 19 01:37:08 2012
From: raidalzobaidi at yahoo.com (Raaed)
Date: Mon, 19 Mar 2012 02:37:08 -0700 (PDT)
Subject: [Hotspot] Modeling 3D IC with Hotspot
In-Reply-To: <CAH8TaYn1gaUXFyC0jh3VCi8mm5KJofdx=pF7+JtZheqgewfsVA@mail.gmail.com>
References: <CAB-vPZMN8KWSR8vDD9Qy7BakN4hjh8OfO4iuh9eDfa9dXRjVUw@mail.gmail.com>
	<CAH8TaY=vRSCxE2+2egc9bCW30worU2W-YeEy7uURwmrcsaQg=w@mail.gmail.com>
	<CAB-vPZNpZSnJeBcy1w9rTa82U+LXMC+6+fsiDsqEoqCMNZfLYg@mail.gmail.com>
	<CAH8TaYn1gaUXFyC0jh3VCi8mm5KJofdx=pF7+JtZheqgewfsVA@mail.gmail.com>
Message-ID: <1332149828.14100.YahooMailNeo@web110010.mail.gq1.yahoo.com>

Re: Dump the steady state grid temperature for every layer in the 3D chip?
Greetings,
?
Any one done this so far can help! otherwise I need to go myself through this and will let you know the results;)
?
Regards, Ra'ed
?


________________________________
From: Simone Corbetta <scorbetta at elet.polimi.it>
To: hotspot at cs.virginia.edu 
Sent: Monday, March 12, 2012 10:46 AM
Subject: Re: [Hotspot] Modeling 3D IC with Hotspot

Dear Arihant,

I believe HotSpot by default dumps out only the last (top) layer of
the 3D grid model. I think you can easily extend the code to dump out
the grid model in every layer. I remember once I came across the
definition of the 'cuboid' data structure in the 'grid_model_vector_t'
struct within 'temperature_grid.h' header. Maybe you can go and take a
look at it :-D

Let's stay in touch about this, I am interested in this, too..

Best reagards

On Sun, Mar 11, 2012 at 4:50 AM, Arihant Jain <arihant at iith.ac.in> wrote:
> Hey Simone...Thanks for your reply.
>
> I m already using the Layer Configuration file(.lcf) to describe various
> layers and thermal conductivity between them and it is also generating the
> temperature of all the blocks across all the ?layer.
>
> What I want is instead of associating one particular temperature to a
> block,it should give me the temperatures of all 64*64 grid so that i can
> plot it using the grid_thermal_map.pl.
>
> Temperature for the all the grid structures can be obtained by using the
> grid_steady_file for a single layer floor plan but the same command cannot
> be used for getting grid temperatures for Multiple layers considered in .lcf
> file.(I tried it)
>
> On Sat, Mar 10, 2012 at 3:26 PM, Simone Corbetta <scorbetta at elet.polimi.it>
> wrote:
>>
>> Hi Arihant,
>>
>> HotSpot has some support for 3D stacking. You should specify a layer
>> configuration file (LCF), as done in the example.lcf example file in
>> the HotSpot release. Major information can be found in
>> the "How To" section on the webpage. However, some tricks are required
>> to model vias.
>>
>> Best regards
>>
>> On Sat, Mar 10, 2012 at 7:55 PM, Arihant Jain <arihant at iith.ac.in> wrote:
>> > Hello,
>> >
>> > I am trying to generate the thermal profile for 2 stacked Alpha 21264
>> > cores
>> > using Grid model.
>> > My Question is:
>> >
>> > How can i generate the temperature corresponding to all the grid
>> > structures(4096 for 64*64) for both the layers and Plot them?
>> >
>> > Hotspot support that for a single layer at a time but in order to
>> > incorporate the heat transfer from layers sitting above and below,we
>> > need to
>> > calculate those grid temperature values?simultaneously.
>> >
>> > Regards
>> > Arihant Jain
>> > 4th Year Student
>> > Bachelor of Technology(Hons)
>> > Department of Electrical Engineering
>> > Indian Institute of Technology,Hyderabad
>> > http://www.iith.ac.in
>> >
>> > Ph.No:+919441919112
>> > Email ID: arihant at iith.ac.in, arihantjain003 at gmail.com
>> >
>> >
>> >
>> > _______________________________________________
>> > HotSpot mailing list
>> > HotSpot at mail.cs.virginia.edu
>> > http://www.cs.virginia.edu/mailman/listinfo/hotspot
>> >
>>
>>
>>
>> --
>> Simone Corbetta, Ph.D. student
>>
>> Politecnico di Milano
>> Dipartimento di Elettronica e Informazione
>> Viale delle Rimembranze di Lambrate 14
>> 20134 Milano, ITALY
>>
>> Office: +39.02.2399.9623
>> Home:?http://home.dei.polimi.it/scorbetta
>> Linkedin:?http://www.linkedin.com/in/thelibrarian
>
>
>
>
> --
>
> Arihant Jain
> 4th Year Student
> Bachelor of Technology(Hons)
> Department of Electrical Engineering
> Indian Institute of Technology,Hyderabad
> http://www.iith.ac.in
>
> Ph.No:+919441919112
> Email ID: arihant at iith.ac.in, arihantjain003 at gmail.com
>
>



-- 
Simone Corbetta, Ph.D. student

Politecnico di Milano
Dipartimento di Elettronica e Informazione
Viale delle Rimembranze di Lambrate 14
20134 Milano, ITALY

Office: +39.02.2399.9623
Home:?http://home.dei.polimi.it/scorbetta
Linkedin:?http://www.linkedin.com/in/thelibrarian

_______________________________________________
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/20120319/aba6b308/attachment.html 

From raidalzobaidi at yahoo.com  Mon Mar 19 02:16:47 2012
From: raidalzobaidi at yahoo.com (Raaed)
Date: Mon, 19 Mar 2012 03:16:47 -0700 (PDT)
Subject: [Hotspot] Modeling 3D IC with Hotspot
In-Reply-To: <1332149828.14100.YahooMailNeo@web110010.mail.gq1.yahoo.com>
References: <CAB-vPZMN8KWSR8vDD9Qy7BakN4hjh8OfO4iuh9eDfa9dXRjVUw@mail.gmail.com>
	<CAH8TaY=vRSCxE2+2egc9bCW30worU2W-YeEy7uURwmrcsaQg=w@mail.gmail.com>
	<CAB-vPZNpZSnJeBcy1w9rTa82U+LXMC+6+fsiDsqEoqCMNZfLYg@mail.gmail.com>
	<CAH8TaYn1gaUXFyC0jh3VCi8mm5KJofdx=pF7+JtZheqgewfsVA@mail.gmail.com>
	<1332149828.14100.YahooMailNeo@web110010.mail.gq1.yahoo.com>
Message-ID: <1332152207.67129.YahooMailNeo@web110011.mail.gq1.yahoo.com>

here you go! it is not elegant solution but its workable one :D. This will write all layers data in the same file. Each set of data separated by three newlines.
?
?
To do it change the two nested loops in the function dump_top_layer_temp_grid in temperature_grid.c file to the following:
?
------------------------------------------
for(k=0;? k < model->n_layers; k++){ /* Ra'ed: to dump grid data of all layers */
?for(i=0;? i < model->rows; i++){
??for(j=0;? j < model->cols; j++){
???fprintf(fp, "%d\t%.2f\n", i*model->cols+j, 
?????/* top layer of the most-recently computed 
????? * steady state temperature?
????? */
?????model->last_steady->cuboid[k][i][j]); 
??? }
???fprintf(fp, "\n");
? }
? fprintf(fp, "\n\n\n");
?}
-------------------------------------------?
?
p.s. dont forget to define the new counter 'k' as integer!
?
Best, Ra'ed
?


________________________________
From: Raaed <raidalzobaidi at yahoo.com>
To: Simone Corbetta <scorbetta at elet.polimi.it>; "hotspot at cs.virginia.edu" <hotspot at cs.virginia.edu> 
Sent: Monday, March 19, 2012 9:37 AM
Subject: Re: [Hotspot] Modeling 3D IC with Hotspot


Re: Dump the steady state grid temperature for every layer in the 3D chip?
Greetings,
?
Any one done this so far can help! otherwise I need to go myself through this and will let you know the results;)
?
Regards, Ra'ed
?


________________________________
From: Simone Corbetta <scorbetta at elet.polimi.it>
To: hotspot at cs.virginia.edu 
Sent: Monday, March 12, 2012 10:46 AM
Subject: Re: [Hotspot] Modeling 3D IC with Hotspot

Dear Arihant,

I believe HotSpot by default dumps out only the last (top) layer of
the 3D grid model. I think you can easily extend the code to dump out
the grid model in every layer. I remember once I came across the
definition of the 'cuboid' data structure in the 'grid_model_vector_t'
struct within 'temperature_grid.h' header. Maybe you can go and take a
look at it :-D

Let's stay in touch about this, I am interested in this, too..

Best reagards

On Sun, Mar 11, 2012 at 4:50 AM, Arihant Jain <arihant at iith.ac.in> wrote:
> Hey Simone...Thanks for your reply.
>
> I m already using the Layer Configuration file(.lcf) to describe various
> layers and thermal conductivity between them and it is also generating the
> temperature of all the blocks across all the ?layer.
>
> What I want is instead of associating one particular temperature to a
> block,it should give me the temperatures of all 64*64 grid so that i can
> plot it using the grid_thermal_map.pl.
>
> Temperature for the all the grid structures can be obtained by using the
> grid_steady_file for a single layer floor plan but the same command cannot
> be used for getting grid temperatures for Multiple layers considered in .lcf
> file.(I tried it)
>
> On Sat, Mar 10, 2012 at 3:26 PM, Simone Corbetta <scorbetta at elet.polimi.it>
> wrote:
>>
>> Hi Arihant,
>>
>> HotSpot has some support for 3D stacking. You should specify a layer
>> configuration file (LCF), as done in the example.lcf example file in
>> the HotSpot release. Major information can be found in
>> the "How To" section on the webpage. However, some tricks are required
>> to model vias.
>>
>> Best regards
>>
>> On Sat, Mar 10, 2012 at 7:55 PM, Arihant Jain <arihant at iith.ac.in> wrote:
>> > Hello,
>> >
>> > I am trying to generate the thermal profile for 2 stacked Alpha 21264
>> > cores
>> > using Grid model.
>> > My Question is:
>> >
>> > How can i generate the temperature corresponding to all the grid
>> > structures(4096 for 64*64) for both the layers and Plot them?
>> >
>> > Hotspot support that for a single layer at a time but in order to
>> > incorporate the heat transfer from layers sitting above and below,we
>> > need to
>> > calculate those grid temperature values?simultaneously.
>> >
>> > Regards
>> > Arihant Jain
>> > 4th Year Student
>> > Bachelor of Technology(Hons)
>> > Department of Electrical Engineering
>> > Indian Institute of Technology,Hyderabad
>> > http://www.iith.ac.in
>> >
>> > Ph.No:+919441919112
>> > Email ID: arihant at iith.ac.in, arihantjain003 at gmail.com
>> >
>> >
>> >
>> > _______________________________________________
>> > HotSpot mailing list
>> > HotSpot at mail.cs.virginia.edu
>> > http://www.cs.virginia.edu/mailman/listinfo/hotspot
>> >
>>
>>
>>
>> --
>> Simone Corbetta, Ph.D. student
>>
>> Politecnico di Milano
>> Dipartimento di Elettronica e Informazione
>> Viale delle Rimembranze di Lambrate 14
>> 20134 Milano, ITALY
>>
>> Office: +39.02.2399.9623
>> Home:?http://home.dei.polimi.it/scorbetta
>> Linkedin:?http://www.linkedin.com/in/thelibrarian
>
>
>
>
> --
>
> Arihant Jain
> 4th Year Student
> Bachelor of Technology(Hons)
> Department of Electrical Engineering
> Indian Institute of Technology,Hyderabad
> http://www.iith.ac.in
>
> Ph.No:+919441919112
> Email ID: arihant at iith.ac.in, arihantjain003 at gmail.com
>
>



-- 
Simone Corbetta, Ph.D. student

Politecnico di Milano
Dipartimento di Elettronica e Informazione
Viale delle Rimembranze di Lambrate 14
20134 Milano, ITALY

Office: +39.02.2399.9623
Home:?http://home.dei.polimi.it/scorbetta
Linkedin:?http://www.linkedin.com/in/thelibrarian

_______________________________________________
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/20120319/7ce8de4d/attachment-0001.html 

From niveneetha at gmail.com  Tue Mar 20 01:24:43 2012
From: niveneetha at gmail.com (nive nivethithakasi)
Date: Tue, 20 Mar 2012 14:54:43 +0530
Subject: [Hotspot] VLSI-Hotspot
Message-ID: <CAB4P2k+zpvZRx8aFfKswoD9TfBZ+K6U=OekeVw8OGi53cO+NcA@mail.gmail.com>

 Hello,
we can do project on nonslicing temperature aware VLSI floorplanning, can
we have a MCNC benchmark circuits input file based on temperature. Anybody
please send the file format to us.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120320/ef3e4ae1/attachment.html 

From wh6p at virginia.edu  Tue Mar 20 09:29:39 2012
From: wh6p at virginia.edu (Wei Huang)
Date: Tue, 20 Mar 2012 12:29:39 -0500
Subject: [Hotspot] Hotspot - Extracting Conductance Values
In-Reply-To: <4F675471.4020409@cs.virginia.edu>
References: <CAFT6x5cOqx03VKek9voVxD5vc6ZiD0==iYuF4ruYVRWymOHZNA@mail.gmail.com>
	<4F675471.4020409@cs.virginia.edu>
Message-ID: <CANYeXBqi70+YuzkoHFnAjCe8zqGCf3Ck4vf+VTYyRvKvZn_yoA@mail.gmail.com>

Hi Spencer,

There are a few routines in HotSpot for calculating and populating R and C
values. For example, populate_package_R() in
temperature.c, populate_R_model_block() in temperature_block.c etc. The
related data structures storing those values can also be found in those
routines. In some cases, you might need to convert from thermal conductance
to thermal resistance, though.

Hope this helps.

-Wei


> -------- Original Message --------
> Subject:        [Hotspot] Hotspot - Extracting Conductance Values
> Date:   Wed, 14 Mar 2012 01:04:51 -0500
> From:   Spencer Millican <smillican at wisc.edu>
> To:     hotspot at mail.cs.virginia.edu
>
>
>
> Hotspot users,
>
> I'm trying to use Hotspot to extract conductance values to use in
> a separate program using a block model to calculate steady state
> temperatures under different power conditions.  I'm certain the values I
> want to extract calculated in temperature_block.c, and I just want to
> add a print statement to give me the values I need and exit.
>
> However, I'm having some troubles finding the values I want.  Can anyone
> help point me to these values?
>
> -Conductance/Resistance between two blocks in the silicon layer
> -Conductance/Resistance between two blocks in the HS layer
> -Conductance/Resistance between a block in the silicon layer and
> its corresponding HS layer block
> -Conductance/Resistance between a block in the HS layer and ambient
>
> Can someone please point me out to the variables containing these
> values?  It would be greatly appreciated.
>
> -Spencer Millican
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120320/ec78abe7/attachment.html 

From chenyun.pan at gmail.com  Mon Mar 26 13:47:49 2012
From: chenyun.pan at gmail.com (Chenyun Pan)
Date: Mon, 26 Mar 2012 17:47:49 -0400
Subject: [Hotspot] About Hotspot
Message-ID: <CAKevoq8=7B8BQzPiThbP0mniaJe+Met5tToh81GNuiWyYon_KA@mail.gmail.com>

Hi,

I used to use HotSpot 2 and now I try to use HotSpot 5. The results of new
Hotspot is much lower than the previous version even though I set the same
parameter like chip thickness. With the new version, if I make the power
density be equal to 100 W/cm2, the temperature is only 70 degree. I also
look at the default floorplan ev6.flp and I ran the simulator, it turned
out to be that the highest temperature in the integer block is only less
than 80 degree. I calculate the power density according to the default
power trace file for that block and it is more than 200 W/cm2, which should
be very high. I am not sure whether it is correct.
Also, the results are almost the same no matter whether I use thermal
interface material (TIM) or not.

Thanks,

Chenyun
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120326/e7182fad/attachment.html 

From wh6p at virginia.edu  Mon Mar 26 14:07:53 2012
From: wh6p at virginia.edu (Wei Huang)
Date: Mon, 26 Mar 2012 17:07:53 -0500
Subject: [Hotspot] About Hotspot
In-Reply-To: <CAKevoq8=7B8BQzPiThbP0mniaJe+Met5tToh81GNuiWyYon_KA@mail.gmail.com>
References: <CAKevoq8=7B8BQzPiThbP0mniaJe+Met5tToh81GNuiWyYon_KA@mail.gmail.com>
Message-ID: <CANYeXBqH-24SJ6_9Fe+ZNzCRQM38VGa7idOiE7Uo3Cz_H-05=Q@mail.gmail.com>

Hi Chenyun,

There have been quite a few changes from HS2.0 to 5.0. Please take a look
at the version history  http://lava.cs.virginia.edu/HotSpot/versions.htm
I would recommend using HS5.0 for more accurate results.

Also, when you mentioned TIM, what did you do to not use TIM. Please
provide more details.

-Wei

On Mon, Mar 26, 2012 at 4:47 PM, Chenyun Pan <chenyun.pan at gmail.com> wrote:

> Hi,
>
> I used to use HotSpot 2 and now I try to use HotSpot 5. The results of new
> Hotspot is much lower than the previous version even though I set the same
> parameter like chip thickness. With the new version, if I make the power
> density be equal to 100 W/cm2, the temperature is only 70 degree. I also
> look at the default floorplan ev6.flp and I ran the simulator, it turned
> out to be that the highest temperature in the integer block is only less
> than 80 degree. I calculate the power density according to the default
> power trace file for that block and it is more than 200 W/cm2, which should
> be very high. I am not sure whether it is correct.
> Also, the results are almost the same no matter whether I use thermal
> interface material (TIM) or not.
>
> Thanks,
>
> Chenyun
>
> _______________________________________________
> 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/20120326/1418cd70/attachment.html 

From skadron at cs.virginia.edu  Mon Mar 26 15:05:39 2012
From: skadron at cs.virginia.edu (Kevin Skadron)
Date: Mon, 26 Mar 2012 19:05:39 -0400
Subject: [Hotspot] About Hotspot
In-Reply-To: <CANYeXBqH-24SJ6_9Fe+ZNzCRQM38VGa7idOiE7Uo3Cz_H-05=Q@mail.gmail.com>
References: <CAKevoq8=7B8BQzPiThbP0mniaJe+Met5tToh81GNuiWyYon_KA@mail.gmail.com>
	<CANYeXBqH-24SJ6_9Fe+ZNzCRQM38VGa7idOiE7Uo3Cz_H-05=Q@mail.gmail.com>
Message-ID: <4F70F643.20301@cs.virginia.edu>

Can you also give us the exact configuration files, floorplan, and power 
trace?
Thanks,
/K

On 3/26/2012 6:07 PM, Wei Huang wrote:
> Hi Chenyun,
>
> There have been quite a few changes from HS2.0 to 5.0. Please take a
> look at the version history
> http://lava.cs.virginia.edu/HotSpot/versions.htm
> I would recommend using HS5.0 for more accurate results.
>
> Also, when you mentioned TIM, what did you do to not use TIM. Please
> provide more details.
>
> -Wei
>
> On Mon, Mar 26, 2012 at 4:47 PM, Chenyun Pan <chenyun.pan at gmail.com
> <mailto:chenyun.pan at gmail.com>> wrote:
>
>     Hi,
>
>     I used to use HotSpot 2 and now I try to use HotSpot 5. The results
>     of new Hotspot is much lower than the previous version even though I
>     set the same parameter like chip thickness. With the new version, if
>     I make the power density be equal to 100 W/cm2, the temperature is
>     only 70 degree. I also look at the default floorplan ev6.flp and I
>     ran the simulator, it turned out to be that the highest temperature
>     in the integer block is only less than 80 degree. I calculate the
>     power density according to the default power trace file for that
>     block and it is more than 200 W/cm2, which should be very high. I am
>     not sure whether it is correct.
>     Also, the results are almost the same no matter whether I use
>     thermal interface material (TIM) or not.
>
>     Thanks,
>
>     Chenyun
>
>     _______________________________________________
>     HotSpot mailing list
>     HotSpot at mail.cs.virginia.edu <mailto: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

From wh6p at virginia.edu  Mon Mar 26 16:39:21 2012
From: wh6p at virginia.edu (Wei Huang)
Date: Mon, 26 Mar 2012 19:39:21 -0500
Subject: [Hotspot] About Hotspot
In-Reply-To: <4F70F643.20301@cs.virginia.edu>
References: <CAKevoq8=7B8BQzPiThbP0mniaJe+Met5tToh81GNuiWyYon_KA@mail.gmail.com>
	<CANYeXBqH-24SJ6_9Fe+ZNzCRQM38VGa7idOiE7Uo3Cz_H-05=Q@mail.gmail.com>
	<4F70F643.20301@cs.virginia.edu>
Message-ID: <CANYeXBpuGZ53dB3ffiaLWZJunQ_XUR-70Y+oBFLLQ6TcGhu3nA@mail.gmail.com>

In addition to Kevin's suggestions
- The lcf file is only used for 3D chips. If a single silicon layer is
modeled, you shouldn't use the lcf file, and TIM layer is included by
default.
- If you are doing transient temperature simulation, there is also a change
in the time constant scaling factor in later HotSpot releases.
- Also, there are two modes of HotSpot: grid model and block model, the two
models' result won't match each other perfectly, because the way the model
is built and the different granularities they target... Inside grid model,
you can also choose to report temperature of center grid cell, hottest grid
cell, or average temperature. All these factors could also lead to
differences in your results.

Wei

On Mon, Mar 26, 2012 at 6:05 PM, Kevin Skadron <skadron at cs.virginia.edu>wrote:

> Can you also give us the exact configuration files, floorplan, and power
> trace?
> Thanks,
> /K
>
> On 3/26/2012 6:07 PM, Wei Huang wrote:
> > Hi Chenyun,
> >
> > There have been quite a few changes from HS2.0 to 5.0. Please take a
> > look at the version history
> > http://lava.cs.virginia.edu/HotSpot/versions.htm
> > I would recommend using HS5.0 for more accurate results.
> >
> > Also, when you mentioned TIM, what did you do to not use TIM. Please
> > provide more details.
> >
> > -Wei
> >
> > On Mon, Mar 26, 2012 at 4:47 PM, Chenyun Pan <chenyun.pan at gmail.com
> > <mailto:chenyun.pan at gmail.com>> wrote:
> >
> >     Hi,
> >
> >     I used to use HotSpot 2 and now I try to use HotSpot 5. The results
> >     of new Hotspot is much lower than the previous version even though I
> >     set the same parameter like chip thickness. With the new version, if
> >     I make the power density be equal to 100 W/cm2, the temperature is
> >     only 70 degree. I also look at the default floorplan ev6.flp and I
> >     ran the simulator, it turned out to be that the highest temperature
> >     in the integer block is only less than 80 degree. I calculate the
> >     power density according to the default power trace file for that
> >     block and it is more than 200 W/cm2, which should be very high. I am
> >     not sure whether it is correct.
> >     Also, the results are almost the same no matter whether I use
> >     thermal interface material (TIM) or not.
> >
> >     Thanks,
> >
> >     Chenyun
> >
> >     _______________________________________________
> >     HotSpot mailing list
> >     HotSpot at mail.cs.virginia.edu <mailto: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
> _______________________________________________
> 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/20120326/9155e920/attachment-0001.html 

From wh6p at virginia.edu  Mon Mar 26 16:57:44 2012
From: wh6p at virginia.edu (Wei Huang)
Date: Mon, 26 Mar 2012 19:57:44 -0500
Subject: [Hotspot] About Hotspot
In-Reply-To: <4F710D89.8060702@cs.virginia.edu>
References: <CAKevoq8g=bdCvF2zH1cJO9PoMoumeAT9oZPc0_MFHF59QK4Lpg@mail.gmail.com>
	<4F710D89.8060702@cs.virginia.edu>
Message-ID: <CANYeXBozqTa9P6AxGYjS3fcWFiYv=XAsqbJrsNNJTsNTUeb6PQ@mail.gmail.com>

Please compare the hotspot.config file, between 5.0 and 2.0. I am sure the
convection thermal resistance is reduced by quite a bit in HS5.0. This is
to keep up with the fact that cooling packages are getting better over the
years. I believe this is the reason you see different temperatures.

-Wei

-------- Original Message --------
> Subject:        Re: [Hotspot] About Hotspot
> Date:   Mon, 26 Mar 2012 20:33:58 -0400
> From:   Chenyun Pan <chenyun.pan at gmail.com>
> To:     skadron at cs.virginia.edu
>
>
>
> Hi Kevin,
>
> I used a very simple floorplan and power trace file as attached. Also, I
> used the default configuration. But I got the steady state temperature
> which is less than 70 degree. I am not sure if it is correct.
>
> Thanks,
>
> Chenyun
>
>
> On Mon, Mar 26, 2012 at 7:05 PM, Kevin Skadron <skadron at cs.virginia.edu
> <mailto:skadron at cs.virginia.**edu <skadron at cs.virginia.edu>>> wrote:
>
>    Can you also give us the exact configuration files, floorplan, and
>    power trace?
>    Thanks,
>    /K
>
>
>    On 3/26/2012 6:07 PM, Wei Huang wrote:
>
>        Hi Chenyun,
>
>        There have been quite a few changes from HS2.0 to 5.0. Please take a
>        look at the version history
>        http://lava.cs.virginia.edu/__**HotSpot/versions.htm<http://lava.cs.virginia.edu/__HotSpot/versions.htm>
>
>        <http://lava.cs.virginia.edu/**HotSpot/versions.htm<http://lava.cs.virginia.edu/HotSpot/versions.htm>
> >
>        I would recommend using HS5.0 for more accurate results.
>
>        Also, when you mentioned TIM, what did you do to not use TIM. Please
>        provide more details.
>
>        -Wei
>
>        On Mon, Mar 26, 2012 at 4:47 PM, Chenyun Pan
>        <chenyun.pan at gmail.com <mailto:chenyun.pan at gmail.com>
>        <mailto:chenyun.pan at gmail.com <mailto:chenyun.pan at gmail.com>**>__>
>
>        wrote:
>
>            Hi,
>
>            I used to use HotSpot 2 and now I try to use HotSpot 5. The
>        results
>            of new Hotspot is much lower than the previous version even
>        though I
>            set the same parameter like chip thickness. With the new
>        version, if
>            I make the power density be equal to 100 W/cm2, the
>        temperature is
>            only 70 degree. I also look at the default floorplan ev6.flp
>        and I
>            ran the simulator, it turned out to be that the highest
>        temperature
>            in the integer block is only less than 80 degree. I
>        calculate the
>            power density according to the default power trace file for that
>            block and it is more than 200 W/cm2, which should be very
>        high. I am
>            not sure whether it is correct.
>            Also, the results are almost the same no matter whether I use
>            thermal interface material (TIM) or not.
>
>            Thanks,
>
>            Chenyun
>
>            ______________________________**___________________
>
>            HotSpot mailing list
>        HotSpot at mail.cs.virginia.edu
>        <mailto:HotSpot at mail.cs.**virginia.edu<HotSpot at mail.cs.virginia.edu>
> >
>        <mailto:HotSpot at mail.cs.__virg**inia.edu <http://virginia.edu>
>        <mailto:HotSpot at mail.cs.**virginia.edu<HotSpot at mail.cs.virginia.edu>
> >>
>        http://www.cs.virginia.edu/__**mailman/listinfo/hotspot<http://www.cs.virginia.edu/__mailman/listinfo/hotspot>
>        <http://www.cs.virginia.edu/**mailman/listinfo/hotspot<http://www.cs.virginia.edu/mailman/listinfo/hotspot>
> >
>
>
>
>
>
>        ______________________________**___________________
>
>        HotSpot mailing list
>        HotSpot at mail.cs.virginia.edu <mailto:HotSpot at mail.cs.**virginia.edu<HotSpot at mail.cs.virginia.edu>
> >
>        http://www.cs.virginia.edu/__**mailman/listinfo/hotspot<http://www.cs.virginia.edu/__mailman/listinfo/hotspot>
>        <http://www.cs.virginia.edu/**mailman/listinfo/hotspot<http://www.cs.virginia.edu/mailman/listinfo/hotspot>
> >
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20120326/7492cc03/attachment.html