Every now and then my computer makes a clicking noise, that sounds
like a hard drive. It made the same noise before, but was accompanied
by problems: restarting and eventually corrupting the OS (XP). I
believe it was a HD making the sound caused by a failing PS, a
diagnosis confirmed by a tech. Since then I replaced not only the PS
with a new 450W, but the mobo/CPU and did an XP repair. When I bought
the new PSU at a locally-owned computer store I asked the salesman
(who seemed to know what he was talking about) was it enough to power
my new system and he said it was, though it was cheap ($40). I asked
why are some PSUs much more expensive and he said the parts are the
same and you're paying for the name.

Everything has been hunky-dory since then, except the clicking noise
has started again, though no restarting or other noticeable problems.
My thinking is it's either one of the HDs going south, so I'm taking
precautions (backing up) or the new PS causing the same problem. What
are you opinions? Should I get a more expensive PSU?

Biostar GeForce 6100-M9 939
AMD 4200+ 2.2G X2
EVGA nVidia 7600GT 256MB PCI-e
2 gigs Kingston RAM
Maxtor 60g SATA
Western Digital 80g IDE (older)
Chaintech Envy24HT-S
450W PSU

sillyputty wrote:
> Every now and then my computer makes a clicking noise, that sounds
> like a hard drive. It made the same noise before, but was accompanied
> by problems: restarting and eventually corrupting the OS (XP). I
> believe it was a HD making the sound caused by a failing PS, a
> diagnosis confirmed by a tech. Since then I replaced not only the PS
> with a new 450W, but the mobo/CPU and did an XP repair. When I bought
> the new PSU at a locally-owned computer store I asked the salesman
> (who seemed to know what he was talking about) was it enough to power
> my new system and he said it was, though it was cheap ($40). I asked
> why are some PSUs much more expensive and he said the parts are the
> same and you're paying for the name.

http://www.jonnyguru.com/PSU/Ghetto/...x/100_1903.jpg

http://www.jonnyguru.com/PSU/VX550W/Dscn3274.jpg


Do these two units appear to use the same internal components?

"sillyputty" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> Every now and then my computer makes a clicking noise, that sounds
> like a hard drive. It made the same noise before, but was accompanied
> by problems: restarting and eventually corrupting the OS (XP). I
> believe it was a HD making the sound caused by a failing PS, a
> diagnosis confirmed by a tech. Since then I replaced not only the PS
> with a new 450W, but the mobo/CPU and did an XP repair. When I bought
> the new PSU at a locally-owned computer store I asked the salesman
> (who seemed to know what he was talking about) was it enough to power
> my new system and he said it was, though it was cheap ($40). I asked
> why are some PSUs much more expensive and he said the parts are the
> same and you're paying for the name.
>
> Everything has been hunky-dory since then, except the clicking noise
> has started again, though no restarting or other noticeable problems.
> My thinking is it's either one of the HDs going south, so I'm taking
> precautions (backing up) or the new PS causing the same problem. What
> are you opinions? Should I get a more expensive PSU?
>

I would go to a different computer shop.......:-). There are huge
differences in quality between low-end and high-end PSUs. This is not a very
technical way of measuring quality, but take a PC Power and Cooling 450w PSU
in one hand and your cheap PSU in the other and you will notice that the PC
P&C unit is probably at least twice as heavy. Most of that weight is a
better transformer and much better cooling solutions. I always get the best
PSU, within reason, for my systems since everything else in the system is
dependent on it operating correctly. About all the cheaper PSUs rate their
units at max startup power and not actual running wattage. IOW, your 450w
may only be a 300w after startup.

Ed

Points taken (even though I couldn't view the photos posted by Larry).
Today I bought an Antec Neopower 500W for $100 and change at Fry's.
The salesman made the same comparison in weights. I remember comparing
stereo receivers the same way. While I was there I looked at some quad
core systems - damn things are already under $1000. By the time I can
afford a new system 8 cores will probably be just as cheap.

I'm wondering how far multi-cores will go 32, 64 cores, etc. or will
there be an entirely new technology?

In alt.comp.hardware.pc-homebuilt sillyputty <(E-Mail Removed)>
wrote:

>Points taken (even though I couldn't view the photos posted by Larry).
>Today I bought an Antec Neopower 500W for $100 and change at Fry's.
>The salesman made the same comparison in weights. I remember comparing
>stereo receivers the same way. While I was there I looked at some quad
>core systems - damn things are already under $1000. By the time I can
>afford a new system 8 cores will probably be just as cheap.
>
>I'm wondering how far multi-cores will go 32, 64 cores, etc. or will
>there be an entirely new technology?
>
Some other paradigm I presume.
Something not quite so obvious right now.
Doubling the number of cores only gains large amounts of speed when
there are large amounts of processes that *can* be shared.
Once each linear process has it's own separate processor, adding more
cores will just slow things down instead of speed things up.

Like all such, there's rapid point-of-diminishing-returns.
I suspect that point will be about four or eight processors at most.
After that, the overhead to handle each extra set of processors will
exceed the speed gained by passing off processes to each one; except in
certain problems where large arrays *can* handle lots of data.

One place that might be is in video processing.
I can easily see *video boards* having 64 or 128 processing units, but
not main CPUs.

Of course, by then, perhaps they'll integrate them all together, for
that very reason.

--
_____
/ ' / ™
,-/-, __ __. ____ /_
(_/ / (_(_/|_/ / <_/ <_

"Frank McCoy" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> In alt.comp.hardware.pc-homebuilt sillyputty <(E-Mail Removed)>
> wrote:
>
>>Points taken (even though I couldn't view the photos posted by Larry).
>>Today I bought an Antec Neopower 500W for $100 and change at Fry's.
>>The salesman made the same comparison in weights. I remember comparing
>>stereo receivers the same way. While I was there I looked at some quad
>>core systems - damn things are already under $1000. By the time I can
>>afford a new system 8 cores will probably be just as cheap.
>>
>>I'm wondering how far multi-cores will go 32, 64 cores, etc. or will
>>there be an entirely new technology?
>>
> Some other paradigm I presume.
> Something not quite so obvious right now.
> Doubling the number of cores only gains large amounts of speed when
> there are large amounts of processes that *can* be shared.
> Once each linear process has it's own separate processor, adding more
> cores will just slow things down instead of speed things up.
>
> Like all such, there's rapid point-of-diminishing-returns.
> I suspect that point will be about four or eight processors at most.
> After that, the overhead to handle each extra set of processors will
> exceed the speed gained by passing off processes to each one; except in
> certain problems where large arrays *can* handle lots of data.
>
> One place that might be is in video processing.
> I can easily see *video boards* having 64 or 128 processing units, but
> not main CPUs.
>
> Of course, by then, perhaps they'll integrate them all together, for
> that very reason.
>

I pretty much agree Frank. I think 8 cores will probably be the end of
the 'Core' technology. We will probably see those in the not so distant
future, maybe a year or so. Intel and Nvidia both have been working on
integrated CPU/GPU technology for some time (I am sure AMD is too). I don't
know if Intel/Nvidia are sharing technology or working independently. I just
assumed they are working on mobile rather than desktop applications, but who
knows? I could see it used in the productivity (business) line of desktops,
but there is still a lot to work out as far as the high-performance market
of desktops. A major die shrink would probably be needed to cater to the
high-performance crowd to compete with current CPU/GPU performance.


Ed