Dell dis get burnt by AMD and Dell then switched to Intel. Whether
that is their motive today is another matter.
Seems a pretty huge leap to be making. What specific events
are you referring to?
Did you look up the temp spec for the K6-II? It's 60C - go look it up
for yourself. That's because of its construction. It has a cover over
the chip which creates an oven effect.
That's ridiculous, it's a heat-spreader like all new CPUs
use... it wasn't fancy nickel-plated copper, but didn't need
to be, because they have far lower heat production.
For a K6-2 to get up to 60C, there had to be a negligence
outside of AMD. It is EXTREMELY easy to keep a K6-2 below
60C, even a heatsink without a fan can easily do that given
a case/airflow-routing designed to accomodate it.
If a K6-2 gets up to 60C, it's not AMD's fault.
However, 60C WILL NOT damage one!
Go read the spec sheets, your assumptions aren't consistent
with the facts.
You are an AMD shill, which is OK because they need people like you to
help them sell their CPUs. But I think you owe it to yourself to look
at the temp specs for the K6-II CPU.
You are MAD?
I have a stack of old K6 boards, they were not problematic
due to the CPUs at all... the Super-7 chipsets were eclipsed
by the BX chipset, but mostly they simply fell into
obsolescence like anything else- because newer technology
was faster.
The irony here is that you are the Intel shill... even
though a VERY large majority of the more technically skilled
builders choose AMD. That doesn't make Intel a "bad" choice
though, indeed during the super 7 era Intel had a better
platform due to things like more polished AGP support,
better memory bandwidth on the chipsets than super-7 had,
and as mentioned previously, the lower cost AMD chips were
more often paired with lower cost boards, power, memory,
etc- to the extent that there was far more than just (which
CPU or chipset) distinguishing different systems.
The Epox motherboard had an alarm but no shutdown. I was available to
do the shutdown. The alarm was set at 60C by default.
"The Epox"?
Well if we want to drift down memory lane, Intel-based
boards of that era didn't have thermal shutdown either. P3
had incorporated a thermal cutoff, which was fine so long as
the core thermal change was slow enough that the on-die
diode could react in time- wasn't hard for that to fail in
same scenarios typically causing AMD chips to fail, ie-
heatsink missing or misinstalled.
Plus, that is most often to
When I replaced the CPU I saw a small spring-loaded thermistor in the
center of the socket that touched the bottom of the CPU chip carrier.
I read an article where some guy installed thermocouples to see what
the difference between the reported and actual temp was. It was off by
as much as 4C for the K6-II.
What's the point though?
A K6-2 will not be damaged by 64C either.
Maybe not stable-crashing at that temp- but so might an
Intel chip (anyone recall a certain P3 1.13GHz?)
I almost went with AMD, but chose the Celeron D because of price &
performance.
No, you chose it for the Intel name.
AMD chips costing no more outperform it.
It was neither the best price or best performance.
It's still quite adequate for SOHO uses, and many overclock
well... but that doesn't make it anything it isn't.
I got the MCI P4 board with the Intel chipset. That is not junk.
Never said it was (or wasn't).
However, MSI has (like a couple other popular manufacturers)
come up with some decent boards and some not-so-decent due
to bugs. Notibly, MSI had some capacitor problems about 3-4
years ago and multiple boards tended to be more picky about
memory. I'm not suggesting other manufacturers didn't have
their own issues though, merely that MSI like the others
make compromises for certain price-points.
My apps are Internet related.
I am not stressing the power supply so I am not concerned. The specs
for the PSU give amperage at each rail from which you can work out the
wattage.
No, you most definitely CANNOT work out the wattage from the
per-rail amperage specs. That is, _IF_ the specs were
honest but if it's a generic, the odds are overwhelming that
the specs on it are for peak momentary values. Ever seen
those $7 speakers rated for 400W, the ones with the 6VA
wall-wart powering them?
I bought a Tripp Lite SOHO UPS (750 VA) and had the tech support
people work out the actual demand I was placing based on the specs for
my PSU (350 W), monitor (2.7 A) - the cable modem and router are
negligible). He told me according to their configurator I was using at
most 300 VA.
That sounds like it could be in the ballpark... depending on
what's in the system of course, and whether monitor and/or a
printer (particularly laser) were on it too.
That is considerably lower than the 850 VA you would come
up with if you used the maximum power consumption for those devices.
That's because manufacturers put the absolute largest power figure on
the label. Typical consumption is 1/2 to 1/3 of the max.
yes, the absolute largest peak surge rating is, IMO, fraud,
because that's not how any other non-PC PSU is rated nor how
the major manufacturers rate them.
For example a Sparkle 300W has a peak closer to 390W, it can
actually sustain 300W. Likewise with other major
manufacturers (including the largest- Delta).
Soon I will be doing a power-off test on the UPS from which I can
ballpark the actual VA consumption. I need to find the circuit breaker
first (Never test a UPS by pulling its power cord or the local network
wire will float above ground and you might fry the router and
everyone's NIC. Use the circuit breaker to disrupt utility power
because it does not interrupt the ground).
I'll make a rough guesstimation- Your system, if using a
basic video card or integrated video and only one HDD,
probably doesn't use over 150W for the box alone. Because
of this, a generic 350W PSU actually capable of only
200-250W would be sufficient to power the system. So long
as you get an acceptible lifespan out of it, everything's
fine. The problem is if/when the cheap fan fails, or the
output caps fail and massive ripple starts wearing out other
parts... or if the power supply has a more serious sudden
failure and doesn't shut off immediately. Hopefully that
won't happen, but it is a gamble and even OEMs, knowing how
much power their systems use, don't opt for generic grade
PSU. Even when Compaq was putting measly 90-150W PSU in
their boxes, they were honest 90-150W PSU, and the build
quality of the parts (beside the smaller transformer) was as
good (and often better than) the modern generic 350W PSU.
I have a generic 350W in the junk pile- never used, it's not
worth the risk.
