chip work unstable in low or high temperature, anyone can throw some light?

our process is 40G, it work fine in room temperature, but it getting unstable in low/high temperature.

any ideal to tracking these kinds issue?
or IR drop, STA can catch this kind of issue?

Regards

Hi, very difficult problem to debug.

I would start by looking at how the failure shows itself. You need to use the process of elimination to attempt to figure out where on the chip things are going bad.

Are the same blocks failing at hot and cold? Or, are you violating hold in one block when cold and violating setup in another block when hot.

Can you pinpoint the logic blocks that are working properly?

How about the logic blocks that are not working?

Is it a setup problem? Try lowering the clock frequency.

Is it a hold problem? Try raising the temperature until the failure goes away.

Is it a glitch hitting edge-sensitive elements? Transport delay simulation might be able to duplicate the problem.

Is it an IR drop problem? Can you tweak VDD to reproduce the problem at room temperature?

Is it a threshold problem? I have not run across this problem, but I think it looks like a setup violation at low temperatures and a hold violation at high temperatures.

hi, jpvSoccer

thanks for your detail reply.
currently, we can not identify which block is failed, so we can do some experiment to identify which block, and reproduce the IR drop condition, and I think the first thing we will check should be the setup violation in hot, and hold violation in cold.

Regards