DfX-NYUAD / Corblivar

Corblivar is a simulated-annealing-based floorplanning suite for 3D ICs
GNU General Public License v3.0
25 stars 12 forks source link

I met trouble after run bash gp.sh #4

Closed wyclike closed 3 months ago

wyclike commented 4 months ago

Firsrtly,I have successfully run the command:../Corblivar n100 /home/wyc/Corblivar/exp/configs/3dies/regular/n100_soft.conf benches/ and it generated some file ,including:.data,.gp,.flp,.ptrace and so on.

Secondly,I run the gp.sh,and it generate some .pdf like :n100_1_thermal.gp.pdf.

But, i found that all xxx_1_HotSpot.gp.pdf couldn't be open because the wrong that ""n100_3_HotSpot.gp" line 54: warning: Cannot find or open file "n100_HotSpot.steady.grid.gp_data.layer_9", i interface that is the hotSpot could run in right way.

Do u have any idea?

jknechtel commented 4 months ago

I'm afraid the last part of your problem description is not very clear. But I think you're trying to view the thermal map obtained from HotSpot? For that, you'd have to run HotSpot first of course. Please see again the comments I have provided recently for https://github.com/DfX-NYUAD/Corblivar/issues/2 as well as https://github.com/DfX-NYUAD/Corblivar/issues/3

Best, Johann

wyclike commented 4 months ago

Thanks,but i still can't generate the right “n100_1_HotSpot.gp.pdf”. After run HosSpot first,i generate :①n100_HotSpot.lcf;②n100_HotSpot.ptrace; ③n100_HotSpot.steady;④n100_HotSpot.steady.grid;⑤:n100_HotSpot.txt; Then ,i run gnuplot n100_1_HotSpot.gp but still failed for :"n100_3_HotSpot.gp" line 55: warning: Cannot find or open file "n100_HotSpot.steady.grid.gp_data.layer_9"

Actually,i found another question is that in the n100_HotSpot.txt,it include some wrong information like:layer_13_Si_passive_3_27:50 295.42 ,and i have only 3 layer ,what is the 13 layer in it?

wyclike commented 4 months ago

i mean that although i have run the HotSpot.sh, but i still haven't the file ""n100_HotSpot.steady.grid.gpdata.layer{number}".I have only the file like "n100_HotSpot.steady.grid" or "n100_HotSpot.steady"

jknechtel commented 4 months ago

Few thoughts here. Are you sure that HotSpot.sh did run correctly? Did you map the HS variable to your local installation of HotSpot? Did you use the tailored HotSpot version from https://github.com/jknechtel/HotSpot ?

As for the layer information in the txt file, these are material layers, not 3D layers, so that number is expected to be larger there.

wyclike commented 4 months ago

Thank a lot!After i change the version of hotspot to the link you provided,it work! cool~ What a solid work you did~

wyclike commented 3 months ago

Hi,i have do something work with this solid work.And i have used some technology to optimize the thermal analysis process,meaning that compared to the power-thurring method, I now have a faster and more accurate technology that can obtain thermal data for each layer. And i want to do it deeper ,such as i want to use the precise thermal data to guide the simulated annealing process to obtain better solutions. Therefore, I would like to take the liberty of consulting your perspective and insights through this channel, and to gain some inspiration from them. Thank you!

jknechtel commented 3 months ago

Great to hear that, and sure, let's do that. Let me know if you have specific questions. Instead of going through this channel of issues, tough, maybe let's use the Wiki function. Or, you could directly implement your changes into a fork and then create some pull request.