On Tue, 24 Oct 2006 04:32:22 -0000, Vek wrote:
>
>
> Well, I think Quackle can heat up a CPU more than most things we do on
> a PC.
>
> Some evidence is on this webpage:
> http://en.wikipedia.org/wiki/CPU_power_dissipation
>
> "When the CPU is idle, it will draw far less than the typical thermal
> power."
>
> I've observed that when I run Quackle my system monitor shows my CPU
> usage at 100%. Nothing else I do brings my CPU usage up to 100% for
> more than a few seconds.
>
> If anyone thinks this is wrong then I wish he would provide some
> documentation to support it.
>
> Kevin Leeds.
Short answer...
Yes, Quackle simulations load the CPU and yes, this generates far more heat
than the typical applications. No, this is not a bad thing, and yes, your
CPU was designed to do exactly this. It's only a problem if your computer
is defective or being mis-used.
Long, boring, winding answer...
What many people don't realize is that computers (and by computers, I'm
referring to the CPU's which run the computer) spend most of their time
doing absolutely nothing. Even when the hard drive light is flailing away
or when you're pegging the cable modem for all its worth, the computer
spends most of its time waiting because there's nothing for it to do. This
is true, of course, because networks and even hard drives are slow in
comparison to the rate at which a processor can typically process
information.
This is an important point in response to one comment Cesar made, so let me
emphasize it. Cesar says that...
> It's probably my hardware, but I'm sure I've run more intensive stuff
> than Quackle on either computer and it's never happened...
In fact, he may have run stuff which more intensively accesses the hard
drive or which uses large amounts of memory, and slows the computer to a
crawl as a result. But just because the computer is running slowly does
not mean the CPU is working hard...it just means that a vital resource is
in contention (perhaps memory, perhaps the hard drive), and programs are
waiting for that resource.
None of this means that processors weren't designed to be fully utilized.
They absolutely are designed to be fully utilized, and any system will work
fine when pushed to the limits assuming the following...
* The system was properly built (i.e. the people building the system aren't
engaging in deceptive practices, and there are no manufacturing defects)
* The system has not been configured to exceed design limitations (most
notably by raising the clock speed beyond design limits)
* The system (and most notably the fans) haven't failed due to wear and
tear.
* The system is properly ventilated (fans not clogged, airways not blocked)
and running in an appropriate environmental temperature (generally you
wouldn't want to exceed an ambient temperature of 100 degrees F).
Quackle simulations are not unique in pushing the CPU in this way. Many
other programs do, although you don't see such programs as often running on
computers in the home. The most typical example I can think of is
Photoshop, which is very computationally intensive when running filters.
But most programs, such as word processors, web browsers, and even
spreadsheets, really don't do that much computation compared to what the
CPU is capable of. As a class of application, only the modern 3D games
come close, although technically those will place a much greater processing
load (and resulting thermal output) on the video hardware rather than the
CPU.
Since so few programs in use by typical users really push the system
thermally, it's entirely possible for a system to appear to function
correctly when, in fact, it has a thermal defect of some sort. A thermal
defect is one of most common hardware failures, so I'm never particularly
surprised when I see a machine suffering from one. Quackle's only
involvement in this is that, since its simulator is so computationally
intensive, it is more likely to expose such a defect than many other
programs.
A bit of trivia...the whole discussion somewhat ironic for me personally.
My day job involves working on a product used by scientists and engineers
called Mathematica. Many Mathematica users are people who expect their
calculations to be done as quickly as possible, and it's not uncommon for
people to complain if some particular calculation *isn't* using 100% of
their CPU. If anything less than 100% is being used, they reason,
Mathematica isn't using as much of the CPU as it could be, and therefore
things aren't fast enough. So I find myself a bit surprised that I'm now
defending 100% CPU utilization of a computationally intensive task.
Sincerely,
John Fultz
[EMAIL PROTECTED]
Champaign Scrabble Club #590
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/quackle/
<*> Your email settings:
Individual Email | Traditional
<*> To change settings online go to:
http://groups.yahoo.com/group/quackle/join
(Yahoo! ID required)
<*> To change settings via email:
mailto:[EMAIL PROTECTED]
mailto:[EMAIL PROTECTED]
<*> To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
