I found a post on hackaday about melting the plastic package with boiling acid. And they opened quite a few chips already and posted some pictures:
http://zeptobars.ru/en/read/how-to-open ... hat-inside

I contacted them and i also asked if they would open donated chips, and they will. I got their address so if anyone has some chips to boil in acid we could donate them and find out what is hiding inside.

Lets try to make a list(if anyone wants to donate) before donating anything and send it to see if they agree with it.

In the 90's, I read an article that AMD took a razor blade to chips, scraped a chip and photographed it. They scraped the chip again and photographed it. They used several chips and took thousands of pictures. This would help you get a picture of what is inside. You don't always need to use acid on a chip. Be careful with acid, because some acid won't come out of your skin. And be careful playing with blades.

Hehe, you was talking to me
So we may discuss it right here.

Welcome to the forum!

What chips did you plan to open next?

Already decapped but not published are Z80, kr580vm80a, MC34063, RTL8111 and ALC888.
Then I am thinking decapping bunch of transistors (FET and BJT) to see differences, and some very simple CMOS logic which could be good example for reverse engineering.

Hi Bars, and welcome!
Do you have the means, and the interest, to strip the metal and re-photograph? I'm thinking, of course, of reversing the transistor netlist, like visual6502.

Cheers
Ed
Edit to add: nice photos by the way! I'd be interested to see some floorplan of the ST chip, to see where the ARM is.
Edit: the ARM M3 core is apparently just 33k gates, and anything from 0.43mm2 in a 0.18u process down to 0.03mm2 in a 40nanometer process. That ST chip is 8.9mm2, so we're looking for something which is 1/20th of the chip, if it's in 0.18u. But probably it's a 90nm process, so we're looking for an 80th of the chip. But then, looking at pages 24 and 25 of this pdf there might be nothing organised to see!

I was stripping metal on kr580ik80a (i8080-compatible) - should be easy on any chips with only 1 metal.

Doing so on more modern chips (like STM32) is near limits of optical microscopy and multiple metal layers would cause lots of pain.

That's fine - all the best chips are single metal!
Did you notice John McMaster seems to be making some progress with opencv?
http://uvicrec.blogspot.co.uk/2012/11/s ... -ehsm.html

Cheers
Ed

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!