Proposed System

[J.Clarke]

On Sun, 12 Oct 2003 13:50:53 GMT

Wrong. It only takes one person for there to *be* a "perceived need".
The number of people undervolting is proportional to the number of
people perceiving the need for it.

You seem to be confusing "need" and "want".

Irrelevent. He switched to another board specifically so he could
undervolt. Which is why I posted it as an example.

However he did not say anywhere that he had a _need_ to undervolt, only
that he wanted to.

Irrelevent. Nobody suggested running a machine for mission-critical
tasks.

Reread the thread.

I'll agree with this, even though its somewhat irrelevent since nobody
suggested such a scenario.

Again, reread the thread.

 

[J.Clarke]

On Sun, 12 Oct 2003 14:01:07 GMT

Alternative was intended to mean... "alternative to what you are
suggesting... alternative to the (current) complexity and cost of
using a completely passive solution using not-off-the-shelf items..."

Simple semantic misunderstanding... lets not make a big deal over it..


In the way that I described in the very next sentance.


No. Its "above average".


No that is not my contention. I don't know how you dreamed that up.

Well, let's see, I suggested that nobody was being done a service by
being led to believe that running a component out of specification was
anything except "getting away with something" and you argued vehemently
that thousands of people were being done a disservice by such a
statement. That would imply that you believe that it is _not_ "getting
away with something" and if it is not then it should be suitable for
mission-critical tasks and there should be no need for the purchaser to
be made aware that it has been done. So which is it?

 

[Thunder9]

How does undervolting a processor make a fan run slower?

Undervolting the processor does not *make* the fan run slower.

If you have a cooler CPU, you can run the fan slower through a variety
of mechanisms.

Regards,
Thunder9

 

[Thunder9]

A few of hobbyists talking about their projects does not constitute a
perceived need to undervolt anymore than a few automobile enthusiasts
their land speed record attempts constitutes a perceived need for
supersonic rocket cars.

Wrong. It only takes one person for there to *be* a "perceived need".
The number of people undervolting is proportional to the number of
people perceiving the need for it.

In any case the guy you linked says up front that the board he started
with was running things considerably _above_ the manufacturer's
specified operating range.

Irrelevent. He switched to another board specifically so he could
undervolt. Which is why I posted it as an example.

Personally I'd try a little bit more powerful fan. The second quietest
fan on the market moves a good deal more air but is still very quiet.


Yes, many of them know what they are doing, and one thing that they are
not doing is using those out-of-spec machines for mission-critical
tasks or recommending that others do so.

Irrelevent. Nobody suggested running a machine for mission-critical
tasks.

Yes, the ones who know what they are doing are aware that there are
consequences and they are also aware that the fact that their machine
posts and runs a few tests without crashing does not mean that it is
ready to be installed as a mission-critical server whose failure would
cost large amounts of money or an engineering-design workstation in
which inaccurate calculations could cost lives.

I'll agree with this, even though its somewhat irrelevent since nobody
suggested such a scenario.

Regards,
Thunder9

 

[Thunder9]

Of course they can--add a couple of more heat pipes, use both sides of
the case instead of just one, . . .


Uh, custom built passive cooling devices _are_ "alternative cooling
solutions".

Alternative was intended to mean... "alternative to what you are
suggesting... alternative to the (current) complexity and cost of
using a completely passive solution using not-off-the-shelf items..."

Simple semantic misunderstanding... lets not make a big deal over it..

In what way?

In the way that I described in the very next sentance.

Reading comprehension a bit lacking?

No. Its "above average".

Is is your contention that running
a CPU outside the manufacturer's specified operating range is a normal
procedure and that a reseller selling machines so constructed without
informing the purchaser is behaving ethically and that such machines are
to be trusted with mission-critical tasks?

No that is not my contention. I don't know how you dreamed that up.

Regards,
Thunder9

 

[Thunder9]

Why should I undervolt when I can achieve my requirements with a nice
slow fan and no undervolting?

There is no reason to undervolt when you can achieve your requirements
with a nice slow fan.

Regards,
Thunder9

 

[Thunder9]

It is customary to put all of your thoughts concerning a particular post
in a single response to that post.

Apologies. I got so lost in all your incorrect, irrelevent remarks
that I lost my place. I should be more careful in the future.

Fine, forget passive coolers. Please explain why you are unable to cool
your 2.4 GHz P4 with an off-the-shelf heat sink and a Papst 8412NGL
without undervolting.

I clearly stated this is my first home built system, and its proposed.
I will attempt to cool it with off-the-shelf items, perhaps the one
you suggest. If that works I'll be happy as a camper because I won't
need to undervolt and risk stability issues as you have pointed out.

Regards,
Thunder9

 

[J.Clarke]

On Sun, 12 Oct 2003 14:09:40 GMT

Apologies. I got so lost in all your incorrect, irrelevent remarks
that I lost my place. I should be more careful in the future.


I clearly stated this is my first home built system, and its proposed.
I will attempt to cool it with off-the-shelf items, perhaps the one
you suggest. If that works I'll be happy as a camper because I won't
need to undervolt and risk stability issues as you have pointed out.

I missed that it was your first home built machine. As a person who has
built many machines and has taught classes in which others have built
their own first machines, may I _strongly_ suggest that you keep things
as simple as possible--get it working with adequate cooling and at
normal voltage before you try to get fancy.

I'm not saying not to experiment with undervolting if it's something
that you want to do, but do not _count_ on it working--it often does,
but there are manufacturing tolerances on semiconductors which affect
their operating margins and so one processor might undervolt fine and
another from the same lot but a different part of the wafer might not
under the same conditions, just as one might overclock fine but another
not. And when running out of spec, unless one knows specifically what
to test one can never be completely sure that the processor performs all
operations properly--this risk is acceptable for many purposes but not
for all.

And I apologize if I seemed to be attacking you, that was not my
intent--I was attacking the fellow who did not seem to understand that
his claims (such as that the processors are _designed_ to run at lower
voltage than specified and that undervolting will result in a useful
increase in the service life of a system) were sufficiently outrageous
to merit a request for justification.

Hang around USENET long enough and you'll probably end up treating
people who won't support their claims the same way that I do.

 

[kony]

And I apologize if I seemed to be attacking you, that was not my
intent--I was attacking the fellow who did not seem to understand that
his claims (such as that the processors are _designed_ to run at lower
voltage than specified and that undervolting will result in a useful
increase in the service life of a system) were sufficiently outrageous
to merit a request for justification.

Hang around USENET long enough and you'll probably end up treating
people who won't support their claims the same way that I do.

Your attitude is exactly why you received supportive evidence. People
who are willing to learn are worth the time, but those who aren't,
aren't. You'll remain ignorant because you have a fixed idea in your
mind that's not based on fact, but argue as if it were. I very seldom
call someone an idiot, but it applies here, because you've never even
bothered to TEST your theories. That is a basic requirement for any
kind of scientific method, instead of voodo superstition.

If you don't know how to do this extensive testing yourself, and your
system isn't CERTIFIED to be suitable for these critical uses you seem
to need, then you are not eliminating risk by running at spec'd
voltage. Instead you are assuming it, instead of proving it to any
extent at all... not necessarily to me, but even to yourself.


Dave

 

[JAD]

the Dahli Llama has spoken!

Your attitude is exactly why you received supportive evidence. People
who are willing to learn are worth the time, but those who aren't,
aren't. You'll remain ignorant because you have a fixed idea in your
mind that's not based on fact, but argue as if it were. I very seldom
call someone an idiot, but it applies here, because you've never even
bothered to TEST your theories. That is a basic requirement for any
kind of scientific method, instead of voodo superstition.

If you don't know how to do this extensive testing yourself, and your
system isn't CERTIFIED to be suitable for these critical uses you seem
to need, then you are not eliminating risk by running at spec'd
voltage. Instead you are assuming it, instead of proving it to any
extent at all... not necessarily to me, but even to yourself.


Dave

 

[J.Clarke]

Your attitude is exactly why you received supportive evidence. People
who are willing to learn are worth the time, but those who aren't,
aren't. You'll remain ignorant because you have a fixed idea in your
mind that's not based on fact, but argue as if it were.

And that "fixed idea" (other than that you are a twit) is what?

I very seldom
call someone an idiot, but it applies here, because you've never even
bothered to TEST your theories.

What theories are those?

That is a basic requirement for any
kind of scientific method, instead of voodo superstition.

Coming from someone who has not posted one single verifiable fact to
support his argument that's a real hoot.

If you don't know how to do this extensive testing yourself, and your
system isn't CERTIFIED to be suitable for these critical uses you seem
to need, then you are not eliminating risk by running at spec'd
voltage. Instead you are assuming it, instead of proving it to any
extent at all... not necessarily to me, but even to yourself.

So how does one "CERTIFY" such a system? And how do you know that your
"extensive testing" has in fact confirmed correct operation? And you
have no idea what I am assuming. The fact that I want you to answer
certain questions does not mean that I do not know the correct answers,
it means that I want to know _your_ answers.

You don't seem to be able to understand that I am not trying to sell a
particular viewpoint, I'm trying to find out if you know what you're
talking about, and it's becoming increasingly clear that you do not.

 

[J.Clarke]

On Mon, 13 Oct 2003 01:54:26 GMT

Wrong.
Dictionary: Need - n. 1. A condition or situation in which something
is required or wanted. Source: The American Heritage® Dictionary of
the English Language, Fourth Edition


For what? I googled on
mission OR critical intitle:"Proposed System"

and the only relevent hits were you.

Try reading it instead of googling for keywords and you'll find that
other terms were used that translate to "mission critical". The thread
is not that long.

 

[J.Clarke]

On Mon, 13 Oct 2003 02:10:06 GMT

Neither. I argued that thousands were being done a service.

So what was the nature of that service?

 

[Thunder9]

You seem to be confusing "need" and "want".

Wrong.
Dictionary: Need - n. 1. A condition or situation in which something
is required or wanted. Source: The American Heritage® Dictionary of
the English Language, Fourth Edition

Again, reread the thread.

For what? I googled on
mission OR critical intitle:"Proposed System"

and the only relevent hits were you.

Regards,
Thunder9

 

[Thunder9]

Well, let's see, I suggested that nobody was being done a service by
being led to believe that running a component out of specification was
anything except "getting away with something" and you argued vehemently
that thousands of people were being done a disservice by such a
statement. That would imply that you believe that it is _not_ "getting
away with something" and if it is not then it should be suitable for
mission-critical tasks and there should be no need for the purchaser to
be made aware that it has been done. So which is it?

Neither. I argued that thousands were being done a service.

Regards,
Thunder9

 

[Thunder9]

I missed that it was your first home built machine. As a person who has
built many machines and has taught classes in which others have built
their own first machines, may I _strongly_ suggest that you keep things
as simple as possible--get it working with adequate cooling and at
normal voltage before you try to get fancy.

I'm not saying not to experiment with undervolting if it's something
that you want to do, but do not _count_ on it working--it often does,
but there are manufacturing tolerances on semiconductors which affect
their operating margins and so one processor might undervolt fine and
another from the same lot but a different part of the wafer might not
under the same conditions, just as one might overclock fine but another
not. And when running out of spec, unless one knows specifically what
to test one can never be completely sure that the processor performs all
operations properly--this risk is acceptable for many purposes but not
for all.

And I apologize if I seemed to be attacking you, that was not my
intent--I was attacking the fellow who did not seem to understand that
his claims (such as that the processors are _designed_ to run at lower
voltage than specified and that undervolting will result in a useful
increase in the service life of a system) were sufficiently outrageous
to merit a request for justification.

Thanks that's good advice.

Hang around USENET long enough and you'll probably end up treating
people who won't support their claims the same way that I do.

I doubt it. I've been on USENET longer than most. And the longer I'm
on it, the more I learn its a waste of time to engage in such
frivolous blather.

Regards,
Thunder9

 

[chrisv]

Well, if he's out to prove that he's a clueless idiot, it hasn't been
a total waste of time...

LOL.
That's exactly what I was thinking... clicked <send> right in the
middle of a sentence.

Are all you clueless, logically-handicapped idiots still in grade
school? No, I'm sure you are all electrical engineers who really
understand how CPU's work. Guffaw.

 

[chrisv]

reality check....first your saying that everyone is going to have those options in their bios 2nd even if they did all they would

For Christ's sake fix your line width. And stop top posting. And
learn how to spell. Maybe then someone will take what you have to say
seriously.

 

[chrisv]

(top posting fixed)

Toady kill filed.....

*plonk* goes the stupid top poster.

 

[JAD]

all you can do is comment on your stupid nazi approach to communicating,
Most people's ignorance is only exceeded by their desire to express it... good job