31 Jul
2000
31 Jul
'00
12:22 p.m.
I have been wondering how difficult it would be to pop
the top off
an ASIC (perhaps one of the DEC gate arrays in my 11/750), photograph
it using the probe station at work, and reverse engineer the circuitry
based on the photographs. A program to recognize individual transistors
wouldn't be too difficult, then generating simple gates (nand,nor,invert),
from them, the high level stuff like registers and busses...
Anybody know of a tool to do something like that?
I don't know of a tool to do that. On a similar vein, though, I do know of
one company that is designing high speed imaging microscopes to observe
switching in high density logic designs. We were talking about difficulty
in debugging FPGA designs when you're not able to route a signal to a pin.
Image how tough it is when you're making a real, non-programmable device.
Then imagine every gate being a test point, because of the few photons
it gives off when it switches, timed to the nanosecond. Not being in the
business I don't know if this is news or not. But it seemed pretty cool
when I heard about it.
I imaging if you got your hands on one of those ($$$$$$) you could probably
reverse engineer just about anything. :)
Eric