Re: Can a computer virus kill the CPU?

Addressing some 6 November myths from Sebastian, Casper, and Todd:

For example, even in 2006 many Intel CPUs died by overheat.

Well how does that happen? Intel CPUs throttle back when hot. Where
is this overheating? Intel thermal protection has been common
knowledge since the 1980s. Why did I mention it? Because I knew some
have so little technical knowledge as to 'take the bait'; 'assume' an
Intel CPU can overheat. No wonder they also 'know' viruses can damage
hardware.

Stupid. Your description also applies to Flash RAMs, which
can cause unintentional states, of course

Insult posted along with what "unintentional states"? Insult
word alone implies he was caught lying. Poster clearly has limited
technical knowledge. Somehow, he assumes some machine code
instructions executed in a particular order (unintentional state) will
cause hardware damage. Funny. Every machine code instruction in every
CPU can be executed in any order ... and not cause hardware damage.
Instructions executed in wrong order can cause a software crash.
Software crash is not hardware damage. Somehow Sebastian can swap two
machine code instructions and cause damage? No further comment on his
ridiculous assumption is required.

Wrong again. Improper connections can certainly cause damage
due to bad electrical bus arbitration.

Well maybe Sebastian would define this bus? He does this - make
claims without supporting facts - or even name the bus. What is that
bus standard? Sebastian, list by name that bus and define that
arbitration sequence that causes hardware failure. You said it can
happen. Therefore you can example such hardware damage. Arbitration
failures cause software crashes. Furthermore, better software does not
crash due to arbitration failures. Sebastian does not define this bus
damage, cannot define the driver family involved, and cannot define the
failed part. But somehow he knows a virus can cause hardware damage.

This is even a known phenomena when hot-swapping
common Flash RAMs for BIOSes.

Apparently Sebastian assumes the lurker is a fool. Hot-swapping has
nothing to do with a virus. And hot-swapping anything inside the
computer is a first year neophyte violation. Apparently Sebastian
wants lurkers to believe hot-swapping has some relevance to viruses.
Hot-swapping of hardware not specifically designed for hot-swapping is
taboo - that much a violation. Sebastian with technical knowledge
would have known that. Did he post to intentionally deceive lurkers,
or is he that technically naive?

This couldn't be more false in the microprocessor realm
where the state of the art requires reinvention of
architecture implementations to achieve speed gains.

Todd should first define this whole new architecture. For example,
Pentium is nothing more than a slow upgrade over decades from the 8086
- which in turn was an upgrade from the 8080. But then Todd, tell me
about your design experience with an 8080 microprocessor. Show me
yours and I will show you mine. I have a few years in this business
- a little hint. There is no major reinvention which is why the same
development tools are used each year - sometimes with upgrades. Each
new microprocessor is but an upgrade from previous developments, which
is why hardware is so reliable and predictable. Which is why the same
testing software and other tools are used on each new 'generation' CPU
using minor upgrades.

So Todd, show us how a virus can change a processor's microcode?
Show us how "dynamic logic" (whatever that is) is manipulated by a
virus? Show us where a "transistors drawing crowbar current from the
supply rails" exists. Please define which logic family shorts out the
power supply because of a software instruction - a virus? Todd, even
hardware drivers are routinely designed to current limit so that no
crowbar exists. A virus will not change that hardware.

Really? The most common cause of power supply failure
seems to be a short in the system (e.g., one cause by a
loose screw).

If true, then a 'computer expert' who built Casper's system bought a
power supply only using two numbers - dollars and watts. Even 30 years
ago, such failures were not acceptable. Intel specs for power supplies
even list a wire gauge (wire size) to test supply outputs shorted with
no damage.

Unfortunately, I see this too often. Too many 'computer experts'
somehow 'know' because they shotgunned (swapped) a power supply.
Person did not learn why that supply failed - did not even perform an
autopsy - and yet somehow knows. Often, a power supply failed when a
human did not know basic functions required of all supplies. Resulting
failure directly traceable to a human who did not even know of
essential supply functions - but somehow just *knew* a short circuit
could cause damage.

Casper demonstrates a common problem. He lost three power supplies
because a speaker had caused a short? Casper - a power supply even 30+
years ago meet this Intel requirement. Anyone with minimal electrical
knowledge would know this: from Section 3.5.2:
The power supply shall be capable of withstanding a
continuous short-circuit to the output without any
damage or overstress to the unit ... The maximum
short circuit energy in any output shall not exceed
240 VA.

Casper demonstrates why some know without first learning basic
knowledge. Casper with basic knowledge could describe fold-back
current limiting. Instead a loose screw somehow cause power supply
damage. Either Casper had no idea why a power supply failed or Casper
purchased using classic 'bean counter' expertise. Yes, one reason for
power supply failure are 'bean counters' pretending to be technically
informed. Common in an America that needs immigrants as engineers
because some Americans are too lazy to first learn basic electricity
concepts.

Those who assume viruses can cause hardware damage have been kind
enough to demonstrate insufficient technical knowledge. This first
example is most damning. A CPU can be damaged if machine code
instructions are executed in a particular order? OK, so I do have more
comments about Sebastian's post. But then I am appalled at
'soundbyte reasoning'. This thread only proves how easily people
were convinced that Saddam had weapons of mass destruction -
speculation somehow becomes fact. Some people just know without first
learning numbers, standards, and technology. Same people tell us that
a virus can cause hardware failure? Welcome to the reason why
Americans are also dying in Iraq.

Virus does not cause computer hardware damage. Some who disagree
were kind enough to demonstrate how 'soundbyte reasoning' works -
assume that viruses cause hardware failure and post it as fact.

.