Which temp are right.

"Phil Weldon" wrote in message
m...
'Dumbo' wrote, in part:
... I just want to know which one is most accurate so I can do a bit of
overclocking without the worry of burning anything out.
_____

It isn't possible to "burn anything out" by overclocking. As has been
pointed out, the CPU will lock up and cool off long before any damage is
done. Moreover, the 'fail-safe' on-CPU-die thermal diode is always
functional. Raising the CPU core voltage too high can cause immediate
destruction of the CPU, but damage from overheating just isn't possible.

Phil Weldon

In addition to Phil's advise, errors and crashes show up long before you
will burn anything up. I have been overclocking since my first desktops
(wife says since dirt) and have never destroyed a cpu that I can remember.
Since the Pentium (maybe Pro??) there has been the protection from
overheating on Intel processors. I think AMD began after the TBirds with the
Athlon/XP series. I have also never seen the useful lifetime of a processor
shortened by overclocking. I intentionally ran this i7 at extremely high
temps mainly to see if it would throttle back or cause errors. In trying to
get to 3.8Ghz, the cpu would not begin throttling until it was at or just
below 100C. When it would throttle, the temps would drop to safer levels. I
could not get it to completely shut down or even cause errors. It would just
slow down a little. I was pushing the vcore a bit and like Phil said, that
is what can kill the processor and I wouldn't advise using voltages that
bring the temps up like that. I always try and read up on what vcore
voltages others have safely used and then go from there. Every processor can
be different and I have seen many that would run stable at much lower
voltages than others. Excessive heat may cause instability, but don't worry
about killing the processor.

Ed

Thanks for your advice. I always imagined that the manufacturers always
said that it shortened the life of CPUs and then too much would cook the
chip. I have been overclocking since my first 80486dx PC which ran as a
DX50 without hitch but I feared that the overheating would burn the CPU out.
My Q6600 was supposed to have run in a P5K Asus board at 3.5GHz so I thought
a sensible 3GHz on an ROG board should be easy. All the posts I have read
keep on about temps and how high they are. If this isn't so vital why do
they all moan that the temp is way too high. Voltage I can understand but
if all they have to worry about is whether or not the CPU will throttle then
why not go and try it and see.
Some people have been saying that my original settings and temp were way too
high but they were at standard voltages and speeds. I had even tried
reducing the voltage to reduce the temp in fear of burning it out the
processor but now I feel like trying upping the voltage til I get it stable
but without going OTT.

Robin

'Dumbo' wrote:
Thanks for your advice. I always imagined that the manufacturers always
said that it shortened the life of CPUs and then too much would cook the
chip. I have been overclocking since my first 80486dx PC which ran as a
DX50 without hitch but I feared that the overheating would burn the CPU
out. My Q6600 was supposed to have run in a P5K Asus board at 3.5GHz so I
thought a sensible 3GHz on an ROG board should be easy. All the posts I
have read keep on about temps and how high they are. If this isn't so
vital why do they all moan that the temp is way too high. Voltage I can
understand but if all they have to worry about is whether or not the CPU
will throttle then why not go and try it and see.
Some people have been saying that my original settings and temp were way
too high but they were at standard voltages and speeds. I had even tried
reducing the voltage to reduce the temp in fear of burning it out the
processor but now I feel like trying upping the voltage til I get it
stable but without going OTT.
_____

Well, manufactures have their own reason for warning against overclocking,
but now, even Intel is making overclocking friendly motherboards. And lots
of people say what they don't know. In fact, most overclockers run CPUs and
LOWER than the average temperature of the same CPU, not overclocked, in the
average big manufacturer unit. The only connection between high CPU core
temperatures and overclocking is that to get a higher than rated CPU speed
it is necessary to make trade-offs. Intel wants to make sure that even in
worst case situations of room environment and computer system ventilation
that the CPU will still operate correctly at certified speeds. So one
trade-off you can make for a higher than certified speed is to run your CPU
at lower than maximum rated temperatures. Usually anything below 70 C will
give you a pretty good overclock. And, of course, if you are willing to
spend extra for water cooling, peltier cooling, phase change cooling, or
liquid nitrogen boil-off cooling, you can get amazing overclocks. Or you
can boost the CPU core voltage a bit to improve the digital signal
waveforms, however, at a certain point, increasing the CPU core voltage is
no longer useful because of the higher temperatures that may result, or,
ultimately, destruction of the CPU if the core voltage is raised too high.
Not to mention that all the CPUs of the same series are made in exactly the
same way in exactly the same fabrication facilities. AFTER the nominally
identical chips are produced and tested the chips are sorted by how high a
clock speed they reach reliably under adverse conditions. THEN the clock
multipliers are set internally to the selling speeds. Or, even set
internally to meet marketing needs even if higher performing CPUs are set to
lower multipliers. Thus it has been, oh, since, in my personal experience,
at least the Pentium 75 and 90 (I was able to run a Pentium 90 very easily
at 120 MHz - multipliers weren't locked until about the Pentium II 300.

Phil Weldon


"Dumbo" wrote in message
...
Thanks for your advice. I always imagined that the manufacturers always
said that it shortened the life of CPUs and then too much would cook the
chip. I have been overclocking since my first 80486dx PC which ran as a
DX50 without hitch but I feared that the overheating would burn the CPU
out. My Q6600 was supposed to have run in a P5K Asus board at 3.5GHz so I
thought a sensible 3GHz on an ROG board should be easy. All the posts I
have read keep on about temps and how high they are. If this isn't so
vital why do they all moan that the temp is way too high. Voltage I can
understand but if all they have to worry about is whether or not the CPU
will throttle then why not go and try it and see.
Some people have been saying that my original settings and temp were way
too high but they were at standard voltages and speeds. I had even tried
reducing the voltage to reduce the temp in fear of burning it out the
processor but now I feel like trying upping the voltage til I get it
stable but without going OTT.

Robin