[mdlug-discuss] [mdlug] [Fwd: [opensuse-offtopic] AndnowtheManchurianmicrochip]

[Ingles, Raymond]

> From: Aaron Kulkis
> >  Precisely my point. Absent evidence, we have no reason to believe
the
> > 'secret proof'.
> 
> The Air Force never said there were aliens...it was some
> nut job creating hoaxes to sell his nutty books.

 I'm aware of that. 

> We're talking off-the-shelf Intel/AMD machines.

 But if there's a problem with them, more-costly options can be pursued.
If someone stands to make money from those more-costly options, you've
got your motive.

> If you put a sniffer on a network, and data is being sent
> out on that network for no reason at all, you KNOW you've
> got a problem some place.

 Evidence that this has actually happened?

> Who knows.

 I'd be a lot more inclined to believe it if I were given some data or
even an outline of a workable design. Neither has so far been
forthcoming.

>  But China has been pursuing cyber-network warfare
> for over 10 years - I first heard about that threat back in
> the mid-1990's, and at the time, I thought it to be a rather
> preposterous idea.  But then I heard about the trojaned-printer
> that was sent to Hussein's government, and the outbreak of
> viruses, and I realized that the security I had taken for
> granted on Unix and VM/CMS was far from common.

 Speaking of specific, one-shot hacks is different from widespread
subversion.

> No, you don't need to make any such avoidance at all.
> Since the machine has been penetrated since before an OS
> was even installed on it, there's no sudden, noticeable
> decrease in performance of the machine.

 *Except* that we're talking about slipping design changes in at the
fabrication level. The people who design the chips test them, and they
know - *very* precisely - what timings to expect. And they do their best
to test *everything* on the chip; they *have* to. You can't do changes
at that sort of scale without detectable deviations. Like I said before,
the designers have every reason to use every bit of silicon to boost
performance. The fabricators *can't* steal silicon without redesigning
*something*, cutting into performance margins that are already cut as
fine as the designers can get away with.

> And besides, Disk bandwidth compared to memory bandwidth
> is low... the whole reason for DMA controllers is so that
> the data can be sent in burst mode, so as to free up
> the memory bus.

 But if you're doing DMA-related work at the same time, you'll see a
major drop in performance. Dang near everything on the system uses DMA
now.

> >  Umm... no. I don't buy it.
> 
> You also buy the idea that lowering taxes improves the economy,
> and increases employment, despite a few hundred years of data
> which shows that it always works without fail.

 I do? Please cite a case where I've ever said that. Seriously - come up
with a cite or retract that.

 Generally speaking, when I see someone resort to ad hominem I conclude
that they feel they're losing the argument.

 Sincerely,

 Ray Ingles                                   (313) 227-2317

  "No federal income tax was assessed before 1913, because
  government didn't require the kind of dough it needs now
         that it's running a concierge business."
 - Bill Maher, "When You Ride Alone You Ride With bin Laden"
The contents of this e-mail are intended for the named addressee only. It contains information that may be confidential. Unless you are the named addressee or an authorized designee, you may not copy or use it, or disclose it to anyone else. If you received it in error please notify us immediately and then destroy it.