[s-t] needle in haystack digest #3

[Robert Walsh]

> Yes.  I wasn't intending to suggest an attack based solely on masks.

Yes - it'd never work.  Electronic Engineers are a strange bunch.  I've
spent way too much of my time in a room full of people staring at mask
plots of various 50M transistor and up CPUS looking for flaws, areas for
optimization, etc.  It's part of every CPU design cycle and the
developers hate it, but it has to be done.  If the mask doesn't match
the circuit you designed, you can spot that kind of thing fairly
quickly.  If it didn't match in a way that actually did something
useful, then it'd leap right out at you as soon as you unfurled the
plot.

It's not just at the design review stage, either.  The fab people are
constantly processing the chip design, performing their own reviews,
staring through microscopes at the wafers: working on alignment issues,
probing test points, stripping layers and taking cross-sections to
examine everything from VIA quality to metalization issues.  Something
out of the ordinary would make itself obvious pretty quickly.

These people live inside their chips.  That's why they sort of work.

Regards,
 Robert.

--=20
Robert Walsh
Amalgamated Durables, Inc.  -  "We don't make the things you buy."
Email: rjwalsh at durables.org


-----------------------------------------------------------