search

thesourcerer8 / StdCellLib

LibreSilicon's Standard Cell Library Generator
https://libresilicon.com/
Apache License 2.0
17 stars 10 forks source link

challenge_submission #2

Open shivanishah269 opened 4 years ago

shivanishah269 commented 4 years ago

Extraction of frequency from log file and converting in order of GHz and then printing both the values.

thesourcerer8 commented 4 years ago

I am getting many lines that say "Check input file again". Is that intentional? The input file looks good to me.

thesourcerer8 commented 4 years ago

"Maximum frquency in GHz is: 174796.0 GHz" Now that would be really cool with CMOS! By the way, there is a typo in frquency.

shivanishah269 commented 4 years ago

I am getting many lines that say "Check input file again". Is that intentional? The input file looks good to me. Sorry for this bug, it was not intentional.

"Maximum frquency in GHz is: 174796.0 GHz" Now that would be really cool with CMOS! By the way, there is a typo in frquency. Ya, this was my silly mistake in conversion.

I have updated the file.

shivanishah269 commented 4 years ago

I have updated challenge 3.

thesourcerer8 commented 4 years ago

I think you are using the wrong rounding function for this application. Please check.

shivanishah269 commented 4 years ago

I think you are using the wrong rounding function for this application. Please check.

Ya correct. We should the exact frequency if we change it will affect other constraints as well.

thesourcerer8 commented 4 years ago

I think the "maximum frequency" could be rounded down, to avoid issues, but it must not be rounded up. I think more than 5 digits of precision is usually not needed, so I would suggest to round it down to 5 digits. (But what if we have an extremely slow circuit that only can do kilohertz? rounding down to 0 would be bad. So if rounding down would result in 0 we should allow arbitrary percision perhaps.

shivanishah269 commented 4 years ago

I think the "maximum frequency" could be rounded down, to avoid issues, but it must not be rounded up. I think more than 5 digits of precision is usually not needed, so I would suggest to round it down to 5 digits. (But what if we have an extremely slow circuit that only can do kilohertz? rounding down to 0 would be bad. So if rounding down would result in 0 we should allow arbitrary percision perhaps.

Ya correct. I have updated.