From devendra.rai at tik.ee.ethz.ch  Fri Oct 15 02:54:00 2010
From: devendra.rai at tik.ee.ethz.ch (Devendra Rai)
Date: Fri, 15 Oct 2010 11:54:00 +0200
Subject: [Hotspot] Questions about Simulator
Message-ID: <AANLkTikc1HWSv8-yQPgb717rRDHMo9tB-U-z_6PT_5NU@mail.gmail.com>

Hello HotSpot,

I am wondering whether hotspot takes into account:

(1). That the thermal conductance is not fixed for any layer: conductance
decreases with temperature
(2). That power dissipated is a function of temperature too.

Also, if this is already incorporated, it will be great if you can point me
to references where I can find quantitative analysis.

Cheers
Devendra Rai
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20101015/5da1cb0c/attachment.html 

From wh6p at virginia.edu  Mon Oct 18 15:52:30 2010
From: wh6p at virginia.edu (Wei Huang)
Date: Mon, 18 Oct 2010 17:52:30 -0500
Subject: [Hotspot] I have a question with hotspot
In-Reply-To: <AANLkTin36dSjdVPOD61nM03qUYpE-Y02aKPcL0YZ_pLu@mail.gmail.com>
References: <AANLkTin36dSjdVPOD61nM03qUYpE-Y02aKPcL0YZ_pLu@mail.gmail.com>
Message-ID: <AANLkTinhfrZ=Q=_REP0ua9P9Fixb1oywoE2djMVAFrfo@mail.gmail.com>

Hi,

You are right to keep exactly the same units for flp, ptrace and init files.
In your case, it might be the EOF character difference between UNIX and DOS
format. I would recommend to get rid of the # char at the end of your .init
file and copy&paste those chars from sample files in the release, then see
if it works.

-Wei

On Mon, Oct 18, 2010 at 5:32 PM, Michael Jo <mike815jo at gmail.com> wrote:

> Good afternoon.
>
> I am using McPAT and Hotspot at the same time.
> I had a result from McPAT showing power consumption and the area of
> each component.
> I am using with design Alpha21364, which can be the same one as ev6 in
> hotspot.
> In order to generate a new floorplan, I used my own description
> file(alpha.desc) and average power file(alphaavg.p) and used the same
> hotspot.config file.
> After generating the new floorplan, "Alpha.flp", I also made
> "Alpha.ptrace" by using the power consumption results from McPAT and
> modified the given gcc.steady into "Alpha.init" file which describes
> the initial temperature of the design.
> I will list these files as below:
>
> Alpha.flp
> #<unitname>     <width>         <height>        <left_x>        <bottom_y>
> TotalCores      0.00588130297   0.01031710944   0.00201803595
> 0.00452196317
> FLD     0.00202651253   0.00411835598   0.00415013715   0.00040360719
> NOCs    0.00213210120   0.00411835598   0.00201803595   0.00040360719
> MCs     0.00172268923   0.00411835598   0.00617664968   0.00040360719
> L2      0.00991737486   0.00040360719   0.00000000000   0.00000000000
> L2_left 0.00201803595   0.01443546541   0.00000000000   0.00040360719
> L2_right        0.00201803595   0.01443546541   0.00789933892
> 0.00040360719
>
>
> Alpha.ptrace
>
> TotalCores      L2      L2_left L2_right        FLD     NOCs    MCs
>
> 19.6269 1.32635 0.051757        0.051757        0.275566        6.17232
> 2.52713
> 22.4623 1.95874 0.075153        0.084231        0.204536        4.21646
> 1.51375
> 29.5927 2.53541 0.098937        0.098937        0.193681
> 0.962317        0.558153
>
>
> Alpha.init
> #modified from gcc.steady
> TotalCores      322.57
> FLD     324.52
> NOCs    325.45
> MCs     323.54
> L2_left 324.47
> L2      324.04
> L2_right        324.78
> #
>
>
>
> However, when I was running Hotspot in order to generate a new ttrace
> file, I had an error message as below:
>
> @:~/HOTSPOT/HotSpot-5.0$ ./hotspot -c hotspot.config -init_file
> gcc.init -f Alpha.flp -p Alpha.ptrace -o Alpha.ttrace
> error: block Icache not found
>
>
> Therefore, I changed gcc.init into my own init_file
>
> Then, I got:
>
> @:~/HOTSPOT/HotSpot-5.0$ ./hotspot -c hotspot.config -init_file
> Alpha.init -f Alpha.flp -p Alpha.ptrace -o Alpha.ttrace
> error: not enough lines in temperature file
>
> In this case, do I need to list all of the componants listed in the
> given "gcc.steady" file?
> If not, what will be the proper solution for this case?
> Please, let me know if there are other things to modify.
> Thanks!
>
> -Michael.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20101018/9e3a676d/attachment-0001.html 

From solidrepellent at gmail.com  Tue Oct 19 05:46:48 2010
From: solidrepellent at gmail.com (Solid Repel)
Date: Tue, 19 Oct 2010 15:46:48 +0300
Subject: [Hotspot] PTscalar
Message-ID: <AANLkTikBoCuQt2+HiCjoDyvKMXOfZ95AORjO=duOY261@mail.gmail.com>

Hello Hotspot,

I am new to hotspot and other simulators. Can any one let me know how to run
PTscalar (basically the command to be run on linux). As far as I can see
there is no "Readme" corresponding to PTscalar in the archive describing
what it can accomplish and how it is to be run. All the readme files talk
about simplescalar. I understand that this is not the forum to talk about
PTscalar but I assume lot many people would be using it in here and would be
able to help.

Thanks and Regards,
Kamesh.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20101019/99726d7b/attachment.html 

From ks4kk at virginia.edu  Wed Oct 20 11:42:32 2010
From: ks4kk at virginia.edu (Karthik Sankaranarayanan)
Date: Wed, 20 Oct 2010 11:42:32 -0700
Subject: [Hotspot] I have a question with hotspot
In-Reply-To: <AANLkTinhfrZ=Q=_REP0ua9P9Fixb1oywoE2djMVAFrfo@mail.gmail.com>
References: <AANLkTin36dSjdVPOD61nM03qUYpE-Y02aKPcL0YZ_pLu@mail.gmail.com>
	<AANLkTinhfrZ=Q=_REP0ua9P9Fixb1oywoE2djMVAFrfo@mail.gmail.com>
Message-ID: <AANLkTim1_n_zpY2urdMNVZEs+QmBPAgmZ=ZPBmxOBUC3@mail.gmail.com>

Hi Michael,

The initial temperature file needs to specify the internal node temperatures
as well. The easiest way to do this is to first run without specifying an
init_file option and dumping out the steady state temperatures to a file
(using -steady_file option) and resue that file as the set of initial
temperatures for the next run - as explained in the HotSpot HOWTO here:
http://lava.cs.virginia.edu/HotSpot/HotSpot-HOWTO.htm

Here is the snippet from section 1 in the above HOWTO:

a) hotspot -c hotspot.config -f ev6.flp -p gcc.ptrace -o gcc.ttrace
-steady_file gcc.steady
b) cp gcc.steady gcc.init
   hotspot -c hotspot.config -init_file gcc.init -f ev6.flp -p gcc.ptrace -o
gcc.ttrace

Thanks,
-karthik

On Mon, Oct 18, 2010 at 3:52 PM, Wei Huang <wh6p at virginia.edu> wrote:

> Hi,
>
> You are right to keep exactly the same units for flp, ptrace and init
> files. In your case, it might be the EOF character difference between UNIX
> and DOS format. I would recommend to get rid of the # char at the end of
> your .init file and copy&paste those chars from sample files in the release,
> then see if it works.
>
> -Wei
>
> On Mon, Oct 18, 2010 at 5:32 PM, Michael Jo <mike815jo at gmail.com> wrote:
>
>> Good afternoon.
>>
>> I am using McPAT and Hotspot at the same time.
>> I had a result from McPAT showing power consumption and the area of
>> each component.
>> I am using with design Alpha21364, which can be the same one as ev6 in
>> hotspot.
>> In order to generate a new floorplan, I used my own description
>> file(alpha.desc) and average power file(alphaavg.p) and used the same
>> hotspot.config file.
>> After generating the new floorplan, "Alpha.flp", I also made
>> "Alpha.ptrace" by using the power consumption results from McPAT and
>> modified the given gcc.steady into "Alpha.init" file which describes
>> the initial temperature of the design.
>> I will list these files as below:
>>
>> Alpha.flp
>> #<unitname>     <width>         <height>        <left_x>        <bottom_y>
>> TotalCores      0.00588130297   0.01031710944   0.00201803595
>> 0.00452196317
>> FLD     0.00202651253   0.00411835598   0.00415013715   0.00040360719
>> NOCs    0.00213210120   0.00411835598   0.00201803595   0.00040360719
>> MCs     0.00172268923   0.00411835598   0.00617664968   0.00040360719
>> L2      0.00991737486   0.00040360719   0.00000000000   0.00000000000
>> L2_left 0.00201803595   0.01443546541   0.00000000000   0.00040360719
>> L2_right        0.00201803595   0.01443546541   0.00789933892
>> 0.00040360719
>>
>>
>> Alpha.ptrace
>>
>> TotalCores      L2      L2_left L2_right        FLD     NOCs    MCs
>>
>> 19.6269 1.32635 0.051757        0.051757        0.275566        6.17232
>> 2.52713
>> 22.4623 1.95874 0.075153        0.084231        0.204536        4.21646
>> 1.51375
>> 29.5927 2.53541 0.098937        0.098937        0.193681
>> 0.962317        0.558153
>>
>>
>> Alpha.init
>> #modified from gcc.steady
>> TotalCores      322.57
>> FLD     324.52
>> NOCs    325.45
>> MCs     323.54
>> L2_left 324.47
>> L2      324.04
>> L2_right        324.78
>> #
>>
>>
>>
>> However, when I was running Hotspot in order to generate a new ttrace
>> file, I had an error message as below:
>>
>> @:~/HOTSPOT/HotSpot-5.0$ ./hotspot -c hotspot.config -init_file
>> gcc.init -f Alpha.flp -p Alpha.ptrace -o Alpha.ttrace
>> error: block Icache not found
>>
>>
>> Therefore, I changed gcc.init into my own init_file
>>
>> Then, I got:
>>
>> @:~/HOTSPOT/HotSpot-5.0$ ./hotspot -c hotspot.config -init_file
>> Alpha.init -f Alpha.flp -p Alpha.ptrace -o Alpha.ttrace
>> error: not enough lines in temperature file
>>
>> In this case, do I need to list all of the componants listed in the
>> given "gcc.steady" file?
>> If not, what will be the proper solution for this case?
>> Please, let me know if there are other things to modify.
>> Thanks!
>>
>> -Michael.
>>
>
>
> _______________________________________________
> 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/20101020/3b53fde2/attachment.html