From mahaboobalishah at gmail.com Sun Dec 4 22:57:15 2011 From: mahaboobalishah at gmail.com (Mahaboob Alisha) Date: Mon, 5 Dec 2011 12:27:15 +0530 Subject: [Hotspot] Cannot find -lmkl_lpack Message-ID: Respected Sir, Thank you for accepting my request. At present i am learning hotspot tool. Now i was encountered with a problem at the Makefile. when i am using make i am getting an error as Cannot find -lmkl_lpack So i can u please help me how can i come out from this error. also specify the flags waht should be used. I am using intel mkl and in windows XP-32 bit. Thans and Regards Mahaboob Alisha M.Tech, Vlsi Design, SENSE, VIT UNIVERSITY. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111205/3776700b/attachment.html From john.demetriou.jd at gmail.com Tue Dec 6 15:19:41 2011 From: john.demetriou.jd at gmail.com (John Demetriou) Date: Wed, 7 Dec 2011 01:19:41 +0200 Subject: [Hotspot] floorplan Message-ID: i have a class project on 3d processor thermal monitorin and i was assigned to use hotspot to do so. i would like to know how it is possible for me to create a floorplan of a 3d processor in a way that hotspot recognizes so i can get results? John Demetriou -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111207/97d48109/attachment.html From mahaboobalishah at gmail.com Wed Dec 7 03:52:12 2011 From: mahaboobalishah at gmail.com (Mahaboob Alisha) Date: Wed, 7 Dec 2011 17:22:12 +0530 Subject: [Hotspot] about HOTSPOT Message-ID: Respect sir, I am using hotspot as my tool in the academic project. Installation has been completed. i have build the hotspot and libhotspot.a library is created. When i using to build a grid model i am using the command "* hotspot -c hotspot.config -f ev6.flp -p gcc.ptrace \ -o gcc.ttrace -steady_file gcc.steady*" so i am getting an error called *"-bash : hotspot: command not found*". can u please help me out so that i can proceed further. THANKS AND REGARDS MAHABOOBALISHA SYED VIT UNIVERSITY VELLORE -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111207/331dc1ff/attachment.html From vadrao at utu.fi Wed Dec 7 04:30:43 2011 From: vadrao at utu.fi (Vaddina Kameswar Rao) Date: Wed, 07 Dec 2011 14:30:43 +0200 Subject: [Hotspot] about HOTSPOT In-Reply-To: References: Message-ID: Hi Alisha, The command should be *./hotspot -c hotspot.config -f ev6.flp -p gcc.ptrace -o gcc.ttrace -steady_file gcc.steady* Cheers, Kamesh. On Wed, Dec 7, 2011 at 1:52 PM, Mahaboob Alisha wrote: > Respect sir, > > I am using hotspot as my tool in the academic project. > Installation has been completed. i have build the hotspot and libhotspot.a > library is created. > > When i using to build a grid model i am using the command > > "* hotspot -c hotspot.config -f ev6.flp -p gcc.ptrace \ -o gcc.ttrace > -steady_file gcc.steady*" > > so i am getting an error called *"-bash : hotspot: command not found*". > > can u please help me out so that i can proceed further. > > THANKS AND REGARDS > MAHABOOBALISHA SYED > VIT UNIVERSITY > VELLORE > > > _______________________________________________ > HotSpot mailing list > HotSpot at mail.cs.virginia.edu > http://www.cs.virginia.edu/mailman/listinfo/hotspot > > -- University of Turku (Turun Yliopisto), Department of Information Technology, Kameswar Rao Vaddina, Joukahaisenkatu 3-5, 20014 Turku, Finland. Web : http://users.utu.fi/vadrao E-mail : vadrao at utu.fi Phone : +358 2 333 8664 Mobile : +358 440992995 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111207/4ab4d027/attachment.html From skadron at cs.virginia.edu Wed Dec 7 06:38:15 2011 From: skadron at cs.virginia.edu (Kevin Skadron) Date: Wed, 07 Dec 2011 09:38:15 -0500 Subject: [Hotspot] 3D floorplan In-Reply-To: References: Message-ID: <4EDF7A57.4040308@cs.virginia.edu> Please see the "How-To" on the HotSpot website, i.e. http://lava.cs.virginia.edu/HotSpot/HotSpot-HOWTO.htm -- let us know if you have any questions. /K On 12/6/11 6:19 PM, John Demetriou wrote: > i have a class project on 3d processor thermal monitorin and i was > assigned to use hotspot to do so. i would like to know how it is > possible for me to create a floorplan of a 3d processor in a way that > hotspot recognizes so i can get results? > > > > > > John Demetriou > > > > > _______________________________________________ > HotSpot mailing list > HotSpot at mail.cs.virginia.edu > http://www.cs.virginia.edu/mailman/listinfo/hotspot From skadron at cs.virginia.edu Wed Dec 7 07:53:45 2011 From: skadron at cs.virginia.edu (Kevin Skadron) Date: Wed, 07 Dec 2011 10:53:45 -0500 Subject: [Hotspot] Cannot find -lmkl_lpack In-Reply-To: References: Message-ID: <4EDF8C09.3020205@cs.virginia.edu> Are you using the makefile for Windows? /K On 12/5/11 1:57 AM, Mahaboob Alisha wrote: > Respected Sir, > > Thank you for accepting my request. > > At present i am learning hotspot tool. Now i was > encountered with a problem at the Makefile. when i am using make i am > getting an error as > > > Cannot find -lmkl_lpack > > So i can u please help me how can i come out from this error. also > specify the flags waht should be used. > > I am using intel mkl and in windows XP-32 bit. > > Thans and Regards > Mahaboob Alisha > M.Tech, Vlsi Design, > SENSE, > VIT UNIVERSITY. > > > _______________________________________________ > HotSpot mailing list > HotSpot at mail.cs.virginia.edu > http://www.cs.virginia.edu/mailman/listinfo/hotspot From mahaboobalishah at gmail.com Fri Dec 9 01:58:41 2011 From: mahaboobalishah at gmail.com (Mahaboob Alisha) Date: Fri, 9 Dec 2011 15:28:41 +0530 Subject: [Hotspot] about hospot Message-ID: Respected sir, Sir i have tried using ./hotspot -c.......... for grid model.I am getting an error as : error while loading shared library : libmkl_rt.so.:cannot open share the object file:no such file or directory. I am already having this library in mkl root in ia32 but its showing this problem can you please help me out. my Make file configuration is specified as below i am using 32-bit # Intel Machines - acceleration with the Intel # Math Kernel Library (MKL) MATHACCEL = intel INCDIR = /home/10mvd0029/intel/composer_xe_2011_sp1.7.256/mkl/include LIBDIR = /home/10mvd0029/intel/composer_xe_2011_sp1.7.256/mkl/lib/ia32 LIBS = -L$(MKLROOT)/lib/ia32 -lmkl_lapack95 -lmkl_rt -lpthread -lm #EXTRAFLAGS = This is what i am using for my makefile. Is there any changes that i have to make here or any problem which is not related to Makefile. please help me out sir. THANKS Mahaboob Alisha -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111209/fddbb5ab/attachment.html From skadron at cs.virginia.edu Fri Dec 9 03:11:01 2011 From: skadron at cs.virginia.edu (Kevin Skadron) Date: Fri, 09 Dec 2011 06:11:01 -0500 Subject: [Hotspot] about hospot In-Reply-To: References: Message-ID: <4EE1ECC5.6080400@cs.virginia.edu> Which compiler are you using? I should point out that the only compiler environments we have tested are those described in the documentation on the "How-To" section of the website, i.e. http://lava.cs.virginia.edu/HotSpot/HotSpot-HOWTO.htm /K On 12/9/2011 4:58 AM, Mahaboob Alisha wrote: > Respected sir, > > Sir i have tried using ./hotspot -c.......... for grid > model.I am getting an error as : error while loading shared library : > libmkl_rt.so.:cannot open share the object file:no such file or directory. > > I am already having this library in mkl root in ia32 but > its showing this problem can you please help me out. > > my Make file configuration is specified as below > > i am using 32-bit > # Intel Machines - acceleration with the Intel > # Math Kernel Library (MKL) > MATHACCEL = intel > INCDIR = /home/10mvd0029/intel/composer_xe_2011_sp1.7.256/mkl/include > LIBDIR = /home/10mvd0029/intel/composer_xe_2011_sp1.7.256/mkl/lib/ia32 > LIBS = -L$(MKLROOT)/lib/ia32 -lmkl_lapack95 -lmkl_rt -lpthread -lm > #EXTRAFLAGS = > This is what i am using for my makefile. Is there any changes that i > have to make here or any problem which is not related to Makefile. > > please help me out sir. > > THANKS > Mahaboob Alisha > > > _______________________________________________ > HotSpot mailing list > HotSpot at mail.cs.virginia.edu > http://www.cs.virginia.edu/mailman/listinfo/hotspot From mahaboobalishah at gmail.com Mon Dec 12 03:58:37 2011 From: mahaboobalishah at gmail.com (Mahaboob Alisha) Date: Mon, 12 Dec 2011 17:28:37 +0530 Subject: [Hotspot] About hotspot Message-ID: i have completed building the hotspot. Also i have done all the steps which are given in HOWTO by using grid model. output.flp has been created. I am using the step 'fig2dev -L ps' . when i use this command i am not getting any thing on my screen just running. when i teminate this its showing as fig2dev -L ps stoped. I am unable to understand wheather it is taking time to complete if so, can you please suggest how much time it will take to convert or else is that i am doing anything wrong here. Thanks Mahaboob Alisha -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111212/cd73d1b5/attachment-0001.html From wh6p at virginia.edu Mon Dec 12 07:34:42 2011 From: wh6p at virginia.edu (Wei Huang) Date: Mon, 12 Dec 2011 09:34:42 -0600 Subject: [Hotspot] About hotspot In-Reply-To: References: Message-ID: Can you check if you have the right viewable pdf floorplan file generated from previous step? can you check if you have fig2dev installed by "which fig2dev"? Can you also check if there is a .ps file generated when fig2dev is running/finished? -Wei On Mon, Dec 12, 2011 at 5:58 AM, Mahaboob Alisha wrote: > i have completed building the hotspot. Also i have done all the steps > which are given in HOWTO by using grid model. output.flp has been created. > I am using the step > 'fig2dev -L ps' . > when i use this command i am not getting any thing on my screen just > running. > when i teminate this its showing as fig2dev -L ps stoped. > > I am unable to understand wheather it is taking time to complete if so, > can you please suggest how much time it will take to convert or else is > that i am doing anything wrong here. > > Thanks > Mahaboob Alisha > > _______________________________________________ > 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/20111212/7e60d4b9/attachment.html From wh6p at virginia.edu Mon Dec 12 08:09:51 2011 From: wh6p at virginia.edu (Wei Huang) Date: Mon, 12 Dec 2011 10:09:51 -0600 Subject: [Hotspot] About hotspot In-Reply-To: References: Message-ID: Hello, I just noticed you didn't use the exact piped instructions as in HOW-TO. If you don't use the pipe, you'll have to provide input file and output file names to fig2dev, otherwise it won't work. I would suggest use exactly what is in the instructions. or if you want to figure out the command format of fig2dev, you are also welcome to do that, too. -Wei On Mon, Dec 12, 2011 at 9:34 AM, Wei Huang wrote: > Can you check if you have the right viewable pdf floorplan file generated > from previous step? > can you check if you have fig2dev installed by "which fig2dev"? > Can you also check if there is a .ps file generated when fig2dev is > running/finished? > > -Wei > > On Mon, Dec 12, 2011 at 5:58 AM, Mahaboob Alisha < > mahaboobalishah at gmail.com> wrote: > >> i have completed building the hotspot. Also i have done all the steps >> which are given in HOWTO by using grid model. output.flp has been created. >> I am using the step >> 'fig2dev -L ps' . >> when i use this command i am not getting any thing on my screen just >> running. >> when i teminate this its showing as fig2dev -L ps stoped. >> >> I am unable to understand wheather it is taking time to complete if so, >> can you please suggest how much time it will take to convert or else is >> that i am doing anything wrong here. >> >> Thanks >> Mahaboob Alisha >> >> _______________________________________________ >> 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/20111212/45c0971b/attachment.html From ks4kk at virginia.edu Mon Dec 12 10:07:50 2011 From: ks4kk at virginia.edu (Karthik Sankaranarayanan) Date: Mon, 12 Dec 2011 10:07:50 -0800 Subject: [Hotspot] about hospot In-Reply-To: References: Message-ID: Hi Mahaboob Alisha, Looks like the path to the library has not been included in LD_LIBRARY_PATH? Doing the following in your .bashrc should fix it: export LD_LIBRARY_PATH=$LD_LIBRARY_PATH: Hope this helps, -karthik On Fri, Dec 9, 2011 at 1:58 AM, Mahaboob Alisha wrote: > Respected sir, > > Sir i have tried using ./hotspot -c.......... for grid > model.I am getting an error as : error while loading shared library : > libmkl_rt.so.:cannot open share the object file:no such file or directory. > > I am already having this library in mkl root in ia32 but its > showing this problem can you please help me out. > > my Make file configuration is specified as below > > i am using 32-bit > # Intel Machines - acceleration with the Intel > # Math Kernel Library (MKL) > MATHACCEL = intel > INCDIR = /home/10mvd0029/intel/composer_xe_2011_sp1.7.256/mkl/include > LIBDIR = /home/10mvd0029/intel/composer_xe_2011_sp1.7.256/mkl/lib/ia32 > LIBS = -L$(MKLROOT)/lib/ia32 -lmkl_lapack95 -lmkl_rt -lpthread -lm > #EXTRAFLAGS = > This is what i am using for my makefile. Is there any changes that i have > to make here or any problem which is not related to Makefile. > > please help me out sir. > > THANKS > Mahaboob Alisha > _______________________________________________ > 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/20111212/eb34e02f/attachment.html From mahaboobalishah at gmail.com Tue Dec 13 02:02:29 2011 From: mahaboobalishah at gmail.com (Mahaboob Alisha) Date: Tue, 13 Dec 2011 15:32:29 +0530 Subject: [Hotspot] about hotfloorplan Message-ID: Respected Sir, I am using ./hotfloorplan -c hotspot.config........................... so i got some warnings as OMP:warning #72:KMP_AFFINITY:affinity only supported for Intel(R) processors. OMP:warning #71:KMP_AFFINITY:affinity only supported, using "none". and i am getting report as step: t: tried acceps reject avg.cost bestcost can you please explan about that report because i am getting rejected numbers also. so if i get rejects is that any problem that makes to the design -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111213/8ac43b74/attachment.html From mahaboobalishah at gmail.com Tue Dec 13 02:06:30 2011 From: mahaboobalishah at gmail.com (Mahaboob Alisha) Date: Tue, 13 Dec 2011 15:36:30 +0530 Subject: [Hotspot] hotfloorplan Message-ID: Respected Sir, I am using ./hotfloorplan -c hotspot.config........................... so i got some warnings as OMP:warning #72:KMP_AFFINITY:affinity only supported for Intel(R) processors. OMP:warning #71:KMP_AFFINITY:affinity only supported, using "none". Does these warnings make any difference for the design? If so how can i fix those. and i am getting report as step: t: tried: acceps: reject: avg: cost: bestcost: can you please explan about that report because i am getting rejected numbers also. so if i get rejects is that any problem that makes to the design -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111213/5d683497/attachment.html From mahaboobalishah at gmail.com Wed Dec 14 20:57:48 2011 From: mahaboobalishah at gmail.com (Mahaboob Alisha) Date: Thu, 15 Dec 2011 10:27:48 +0530 Subject: [Hotspot] ABOUT hotfloorplan Message-ID: Respected Sir, I am using ./hotfloorplan -c hotspot.config........................... so i got some warnings as OMP:warning #72:KMP_AFFINITY:affinity only supported for Intel(R) processors. OMP:warning #71:KMP_AFFINITY:affinity only supported, using "none". Does these warnings make any difference for the design? If so how can i fix those. and i am getting report as step: t: tried: acceps: reject: avg: cost: bestcost: can you please explan about that report because i am getting rejected numbers also. so if i get rejected tries is that any problem that makes to the design -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111215/977e58cc/attachment.html From mahaboobalishah at gmail.com Tue Dec 20 01:52:20 2011 From: mahaboobalishah at gmail.com (Mahaboob Alisha) Date: Tue, 20 Dec 2011 15:22:20 +0530 Subject: [Hotspot] About hotfloorplan Message-ID: Respected Sir, I am using ./hotfloorplan -c hotspot.config........................... so i got some warnings as OMP:warning #72:KMP_AFFINITY:affinity only supported for Intel(R) processors. OMP:warning #71:KMP_AFFINITY:affinity only supported, using "none". Does these warnings make any difference for the design? If so how can i fix those. and i am getting report as step: t: tried: acceps: reject: avg: cost: bestcost: can you please explan about that report because i am getting rejected numbers also. so if i get rejects is that any problem that makes to the design -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20111220/05d2fcc7/attachment-0001.html From ks4kk at virginia.edu Thu Dec 22 06:14:03 2011 From: ks4kk at virginia.edu (Karthik Sankaranarayanan) Date: Thu, 22 Dec 2011 06:14:03 -0800 Subject: [Hotspot] about hospot In-Reply-To: References: Message-ID: The "inordinate floorplan size" error occurs when the length or width of the die is larger than that of the heat spreader / heat sink. Hope this helps, -karthik On Tue, Dec 20, 2011 at 9:00 PM, Mahaboob Alisha wrote: > Respected Sir, > > Thanks for your reply. So, though there are rejects it > wont effect my design. Thank u. Sir i have given some random floorplan in > ev6.flp to just know the difference to my knowledge but i have got an error > as inordinate floorplan size. Is there any boundaries that i have to take > while i am giving the floorplan or if there is any other that i am doing > mistake. can u please help me out sir. > > Thanks and Regards > Mahaboob Alisha Syed > VIT University > On Wed, Dec 21, 2011 at 12:21 AM, Karthik Sankaranarayanan < > ks4kk at virginia.edu> wrote: > >> Rejects are part of the normal process in simulated annealing. Early on, >> the annealing process accepts all moves - irrespective of whether it lowers >> the objective function or not. As time progresses, more sub-optimal moves >> are rejected, thereby converging towards a low-cost solution. >> >> Hope this helps. >> >> -karthik >> >> >> >> On Tue, Dec 20, 2011 at 1:51 AM, Mahaboob Alisha < >> mahaboobalishah at gmail.com> wrote: >> >>> Respected Sir, >>> >>> I am using >>> ./hotfloorplan -c hotspot.config........................... >>> so i got some warnings as >>> OMP:warning #72:KMP_AFFINITY:affinity only supported for Intel(R) >>> processors. >>> OMP:warning #71:KMP_AFFINITY:affinity only supported, using "none". >>> Does these warnings make any difference for the design? >>> If so how can i fix those. >>> >>> and i am getting report as >>> step: t: tried: acceps: reject: >>> avg: cost: bestcost: >>> >>> can you please explan about that report because i am getting rejected >>> numbers also. so if i get rejects is that any problem that makes to the >>> design >>> >>> >>> >>> On Thu, Dec 15, 2011 at 10:25 AM, Mahaboob Alisha < >>> mahaboobalishah at gmail.com> wrote: >>> >>>> Thanks for your reply sir. I have changed that environmental library >>>> path LD_LIBRARY_PATH. >>>> Now its working >>>> >>>> On Mon, Dec 12, 2011 at 11:37 PM, Karthik Sankaranarayanan < >>>> ks4kk at virginia.edu> wrote: >>>> >>>>> Hi Mahaboob Alisha, >>>>> >>>>> Looks like the path to the library has not been included in >>>>> LD_LIBRARY_PATH? Doing the following in your .bashrc should fix it: >>>>> >>>>> export LD_LIBRARY_PATH=$LD_LIBRARY_PATH: >>>>> >>>>> Hope this helps, >>>>> -karthik >>>>> >>>>> On Fri, Dec 9, 2011 at 1:58 AM, Mahaboob Alisha < >>>>> mahaboobalishah at gmail.com> wrote: >>>>> >>>>>> Respected sir, >>>>>> >>>>>> Sir i have tried using ./hotspot -c.......... for >>>>>> grid model.I am getting an error as : error while loading shared library : >>>>>> libmkl_rt.so.:cannot open share the object file:no such file or directory. >>>>>> >>>>>> I am already having this library in mkl root in ia32 >>>>>> but its showing this problem can you please help me out. >>>>>> >>>>>> my Make file configuration is specified as below >>>>>> >>>>>> i am using 32-bit >>>>>> # Intel Machines - acceleration with the Intel >>>>>> # Math Kernel Library (MKL) >>>>>> MATHACCEL = intel >>>>>> INCDIR = /home/10mvd0029/intel/composer_xe_2011_sp1.7.256/mkl/include >>>>>> LIBDIR = >>>>>> /home/10mvd0029/intel/composer_xe_2011_sp1.7.256/mkl/lib/ia32 >>>>>> LIBS = -L$(MKLROOT)/lib/ia32 -lmkl_lapack95 -lmkl_rt -lpthread -lm >>>>>> #EXTRAFLAGS = >>>>>> This is what i am using for my makefile. Is there any changes that i >>>>>> have to make here or any problem which is not related to Makefile. >>>>>> >>>>>> please help me out sir. >>>>>> >>>>>> THANKS >>>>>> Mahaboob Alisha >>>>>> _______________________________________________ >>>>>> 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/20111222/7970df4e/attachment.html From stasot at microlab.ntua.gr Thu Dec 29 02:26:13 2011 From: stasot at microlab.ntua.gr (Efstathios Sotiriou - Xanthopoulos) Date: Thu, 29 Dec 2011 12:26:13 +0200 Subject: [Hotspot] Processing subareas with hotspot? Message-ID: <1325154373.20228.4.camel@CELINE> Dear all, I would like to ask if we can process a subarea of a floorplan using hotspot, instead of the whole floorplan (suppose that we have information about the steady temperatures of the blocks of the remaining area) Thank you Efstathios Sotiriou-Xanthopoulos National Technical University of Athens From ks4kk at virginia.edu Thu Dec 29 17:08:04 2011 From: ks4kk at virginia.edu (Karthik Sankaranarayanan) Date: Thu, 29 Dec 2011 17:08:04 -0800 Subject: [Hotspot] Processing subareas with hotspot? In-Reply-To: <1325154373.20228.4.camel@CELINE> References: <1325154373.20228.4.camel@CELINE> Message-ID: Hi Efstathios, Component power numbers are required for HotSpot to solve for temperature. That said, if you have steady state thermal information, you could use HotSpot to estimate the power numbers. Please see the following publications for details: [1] Z Qi; B.H. Meyer; W. Huang; R.J. Ribando; K Skadron; M.R Stan; "Temperature-to-power mapping," *ICCD 2010 *, pp.384-389, 3-6 Oct. 2010 http://www.cs.virginia.edu/~skadron/Papers/qi_t2p_iccd10.pdf [2] H. F. Hamann, A. Weger, J. A. Lacey, Z. Hu, P. Bose, E. Cohen, and J. Wakil, ?Hotspot-limited microprcessors: Direct temperature and power distribution measurements,? IEEE J. Solid-State Circuits, vol. 1, no. 42, pp. 56?65, Dec 2007. [3] F. J. Mesa-Martinez, J. Nayfach-Battilana, and J. Renau, ?Power model validation through thermal measurements,? International Symposium on Computer Architecture, vol. 35, no. 2, pp. 302?311, 2007. [4] X. Wang, A. Shakouri, S. Farsiu, and P. Milanfar, ?Extraction of power dissipation profile in an IC chip from temperature map,? in Annual Thermal Measurement, Modeling and Management Symposium, 2007, pp. 51?56. Hope this helps, -karthik On Thu, Dec 29, 2011 at 2:26 AM, Efstathios Sotiriou - Xanthopoulos < stasot at microlab.ntua.gr> wrote: > Dear all, > > I would like to ask if we can process a subarea of a floorplan using > hotspot, instead of the whole floorplan (suppose that we have > information about the steady temperatures of the blocks of the remaining > area) > > Thank you > Efstathios Sotiriou-Xanthopoulos > National Technical University of Athens > > _______________________________________________ > 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/20111229/c1139e6c/attachment.html From stasot at microlab.ntua.gr Fri Dec 30 02:36:11 2011 From: stasot at microlab.ntua.gr (Efstathios Sotiriou - Xanthopoulos) Date: Fri, 30 Dec 2011 12:36:11 +0200 Subject: [Hotspot] Processing subareas with hotspot? In-Reply-To: References: <1325154373.20228.4.camel@CELINE> Message-ID: <1325241371.22856.4.camel@CELINE> Dear Karthik Thank you for your response, however I do not want to calculate the power of the units of the floorplan. Actually, I would like to pose a scenario where there is an incomplete thermal map (with some units only). To find the remaining temperatures, the simple solution is to run hotspot for the whole floorplan, but this may take some time. To accelerate the temperature estimation, is there any functionality provided by the hotspot to calculate the temperature for the missing part only? Thank you Efstathios On Thu, 2011-12-29 at 17:08 -0800, Karthik Sankaranarayanan wrote: > Hi Efstathios, > > Component power numbers are required for HotSpot to solve for > temperature. That said, if you have steady state thermal information, > you could use HotSpot to estimate the power numbers. Please see the > following publications for details: > > > [1] Z Qi; B.H. Meyer; W. Huang; R.J. Ribando; K Skadron; M.R Stan; > "Temperature-to-power > mapping," ICCD 2010 , pp.384-389, 3-6 Oct. 2010 > http://www.cs.virginia.edu/~skadron/Papers/qi_t2p_iccd10.pdf > > [2] H. F. Hamann, A. Weger, J. A. Lacey, Z. Hu, P. Bose, E. Cohen, > and J. Wakil, ?Hotspot-limited microprcessors: Direct temperature and > power distribution measurements,? IEEE J. Solid-State Circuits, vol. > 1, > no. 42, pp. 56?65, Dec 2007. > > [3] F. J. Mesa-Martinez, J. Nayfach-Battilana, and J. Renau, ?Power > model > validation through thermal measurements,? International Symposium > on Computer Architecture, vol. 35, no. 2, pp. 302?311, 2007. > > [4] X. Wang, A. Shakouri, S. Farsiu, and P. Milanfar, ?Extraction of > power > dissipation profile in an IC chip from temperature map,? in Annual > Thermal Measurement, Modeling and Management Symposium, 2007, > pp. 51?56. > > Hope this helps, > -karthik > > > On Thu, Dec 29, 2011 at 2:26 AM, Efstathios Sotiriou - Xanthopoulos > wrote: > Dear all, > > I would like to ask if we can process a subarea of a floorplan > using > hotspot, instead of the whole floorplan (suppose that we have > information about the steady temperatures of the blocks of the > remaining > area) > > Thank you > Efstathios Sotiriou-Xanthopoulos > National Technical University of Athens > > _______________________________________________ > HotSpot mailing list > HotSpot at mail.cs.virginia.edu > http://www.cs.virginia.edu/mailman/listinfo/hotspot > From skadron at cs.virginia.edu Fri Dec 30 11:04:24 2011 From: skadron at cs.virginia.edu (Kevin Skadron) Date: Fri, 30 Dec 2011 14:04:24 -0500 Subject: [Hotspot] Processing subareas with hotspot? In-Reply-To: <1325241371.22856.4.camel@CELINE> References: <1325154373.20228.4.camel@CELINE> <1325241371.22856.4.camel@CELINE> Message-ID: <4EFE0B38.8@cs.virginia.edu> That would be a useful feature, but unfortunately, that's not something we support yet. /K On 12/30/11 5:36 AM, Efstathios Sotiriou - Xanthopoulos wrote: > Dear Karthik > > Thank you for your response, however I do not want to calculate the > power of the units of the floorplan. Actually, I would like to pose a > scenario where there is an incomplete thermal map (with some units > only). To find the remaining temperatures, the simple solution is to run > hotspot for the whole floorplan, but this may take some time. To > accelerate the temperature estimation, is there any functionality > provided by the hotspot to calculate the temperature for the missing > part only? > > Thank you > Efstathios > > > > On Thu, 2011-12-29 at 17:08 -0800, Karthik Sankaranarayanan wrote: >> Hi Efstathios, >> >> Component power numbers are required for HotSpot to solve for >> temperature. That said, if you have steady state thermal information, >> you could use HotSpot to estimate the power numbers. Please see the >> following publications for details: >> >> >> [1] Z Qi; B.H. Meyer; W. Huang; R.J. Ribando; K Skadron; M.R Stan; >> "Temperature-to-power >> mapping," ICCD 2010 , pp.384-389, 3-6 Oct. 2010 >> http://www.cs.virginia.edu/~skadron/Papers/qi_t2p_iccd10.pdf >> >> [2] H. F. Hamann, A. Weger, J. A. Lacey, Z. Hu, P. Bose, E. Cohen, >> and J. Wakil, ?Hotspot-limited microprcessors: Direct temperature and >> power distribution measurements,? IEEE J. Solid-State Circuits, vol. >> 1, >> no. 42, pp. 56?65, Dec 2007. >> >> [3] F. J. Mesa-Martinez, J. Nayfach-Battilana, and J. Renau, ?Power >> model >> validation through thermal measurements,? International Symposium >> on Computer Architecture, vol. 35, no. 2, pp. 302?311, 2007. >> >> [4] X. Wang, A. Shakouri, S. Farsiu, and P. Milanfar, ?Extraction of >> power >> dissipation profile in an IC chip from temperature map,? in Annual >> Thermal Measurement, Modeling and Management Symposium, 2007, >> pp. 51?56. >> >> Hope this helps, >> -karthik >> >> >> On Thu, Dec 29, 2011 at 2:26 AM, Efstathios Sotiriou - Xanthopoulos >> wrote: >> Dear all, >> >> I would like to ask if we can process a subarea of a floorplan >> using >> hotspot, instead of the whole floorplan (suppose that we have >> information about the steady temperatures of the blocks of the >> remaining >> area) >> >> Thank you >> Efstathios Sotiriou-Xanthopoulos >> National Technical University of Athens >> >> _______________________________________________ >> 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