Please answer me -- overclock but undervolt -- safe? How much?

Do you have a question? Post it now! No Registration Necessary.  Now with pictures!

Threaded View
Hi:

Please answer my questions below. Please don't neglect this message.

Is it possible to significantly decrease the heat generated by a  
"Pentium 4 570J/670 3.8 GHz processor" by using extreme undervolting  
without underclocking? How far can undervolting go before it causes  
trouble? What problems can result from undervolting? Can it damage the  
CPU or other parts of the PC?

Also, can these processors be overclocked to 4 GHz while still being  
undervolted to the max tolerable? Does overclocking decrease the extent  
to which undervolting can be safely performed?

In addition, what is the max that these CPUs can be overclocked without  
damaging them or decreasing their lifetimes?

No offense but please respond with reasonable answers & keep out the  
jokes, off-topic nonsense, taunts, insults, and trivializations. I am  
really interested in this.

If I am not answered properly I might go insane and do something  
everyone -- including me -- will eternally regret.


Thanks,

Radium

Re: Please answer me -- overclock but undervolt -- safe? How much?


Quoted text here. Click to load it
<snip>

Why would you want to do this? Just vent your case adequately and choose an  
appropriate hs/f for the unit. I've seen reviews where a 670 did just under  
4.4Ghz with stock voltage and a Zalman 7700 cooler. And they were perfectly  
happy with the temp of the CPU both stock and o/c. If you must, reduce the  
vcore incrementally and test for errors. Part of the huge power consumption  
of the 600 series is the 2mB of on-die L2 cache. This unit consumes over  
160W under full load.

http://www.xbitlabs.com/articles/cpu/display/pentium4-670_3.html



Re: Please answer me -- overclock but undervolt -- safe? How much?




Quoted text here. Click to load it

I have a 630 (EM64T_3ghz) that has been running at 3.6ghz for 2yrs or more
at stock vcore very happily, but any lowering of the vcore causes immediate
instability. These processors are "on the edge" so to speak from the get go,
so expect to underclock if you reduce the vcore. They also tend to run a lot
hotter than today's Core 2/Quad processors so good cooling and airflow is a
must.


Ed



Re: Overclock but undervolt -- safe? How much?



'Radium' wrote:
| Is it possible to significantly decrease the heat generated by "Pentium
| 4 570J 3.8 GHz processor" by using extreme undervolting without
| underclocking? How far can undervolting go before it causes trouble?
| What problems can result from undervolting? Can it damage the CPU or
| other parts of the PC?
|
| Also, can this processor be overclocked to 4 GHz while still being
| undervolted to the max tolerable? Does overclocking decrease the extent
| to which undervolting can be safely performed?
|
| In addition, what is the max that this CPU can be overclocked without
| damaging it or decreasing its lifetime?
_____

Overclocking is not going to damage an Intel x86 CPU before it becomes
obsolete (your Pentium 4 570J 3.8 GHz CPU already is.)  On the other hand,
raising the core voltage too high can immediately destroy an Intel x86 CPU.
Limiting the core voltage to no more than +10% should give a good safety
margin; more than that is possible (there are many accounts of successful
overclocks with a higher core voltage increase, but there are no guarantees.

There is no such thing as 'extreme undervolting.'  There is a cutoff voltage
at which a transistor will no longer operate.  Reducing the core voltage
will not harm a CPU, but you will rapidly reach a voltage at which the CPU
will just not operate.  The CPU will not be damaged, nor will any other part
of the system.  Restoring the core voltage to normal will set everything
aright.

Read Phil's answer carefully, and also other posts in this newsgroup.  You
will benefit from a better grasp of what overclocking entails.  Basically,
just trading off voltage and heat safety margins for higher clock speeds.
Slowly and methodically is the way to go.

Phil Weldon

| Hi:
|
| Is it possible to significantly decrease the heat generated by "Pentium
| 4 570J 3.8 GHz processor" by using extreme undervolting without
| underclocking? How far can undervolting go before it causes trouble?
| What problems can result from undervolting? Can it damage the CPU or
| other parts of the PC?
|
| Also, can this processor be overclocked to 4 GHz while still being
| undervolted to the max tolerable? Does overclocking decrease the extent
| to which undervolting can be safely performed?
|
| In addition, what is the max that this CPU can be overclocked without
| damaging it or decreasing its lifetime?
|
|
| Thanks,
|
| Radium



Re: Overclock but undervolt -- safe? How much?




Quoted text here. Click to load it

In almost every case, NO.

Quoted text here. Click to load it

You can try it........ That processor is pretty much set to operate on it's
default Vcore and lowering it will probably make it unstable even at it's
default speed.

Quoted text here. Click to load it

    The best you could probably do is to find the highest possible freq that
the processor will do at default Vcore, which will probably not be very much
higher. What folks are trying to tell you is that the two processors you
mention are not the most efficient and not great OCers without voltage
INCREASES. They are very marginal even at default speeds and Vcore. Those
processors were Intel's last go-round at using raw speed rather than better
efficiency (operations per cycle) like the Core 2 Duos and Quads. An E6600
at 2.4Ghz outperforms my EM64T 630 OC'd to 3.6Ghz by almost double in any
benchmark. The Q6600 does even better, even in apps that are not optimized
for it. Almost everyone in this group are very knowledgeable and are really
trying to help you out. You just have to listen to what they are telling
you........Regards



Ed



Re: Overclock but undervolt -- safe? How much?



'Radium' wrote, in part:
| It's true that CPUs with lower clock rates can be more efficient than
| CPUs with higher clock rates if the former uses more bits-per-cycle than
| the latter.
|
| As I've recently found, the 4.7 GHz IBM Power6 holds the current record
| for highest frequency CPU:
|
| http://www.spec.org/cpu2006/results/res2007q3/cpu2006-20070708-01383.html
|
| How much does the 4.7 GHz IBM Power6 cost?
|
| Is the 4.7 GHz IBM Power6 available today? If so, where can I find it? I
| live in Southern California in Diamond Bar. Any store close to me that
| carries this processor?
_____

Do some research on your own.  You are just asking a string of questions
that really have no relationship to overclocking.

To repeat, overclocking consist of swapping temperature and voltage safety
margins for higher clock speeds.
Overclocking is operating a CPU at a higher speed than that marked or
specified by the manufacturer.

| It's true that CPUs with lower clock rates can be more efficient than
| CPUs with higher clock rates if the former uses more bits-per-cycle than
| the latter.

 Efficiency, what is efficiency?  Efficiency can be instructions per second,
or instructions per watt, data processing accomplished per watt or second,
instructions per cycle, data processed per cycle.

| How much does the 4.7 GHz IBM Power6 cost?
|
| Is the 4.7 GHz IBM Power6 available today? If so, where can I find it? I
| live in Southern California in Diamond Bar. Any store close to me that
| carries this processor?

Nobody here knows or cares, and you can't afford it.  If you had such a CPU,
what software would you run?

| What is the least expensive motherboard that is fully-compatible with
| this CPU?

See above comment.

| What is the least expensive HDD that is fully-compatible with this CPU?

Hard drives have absolutely NOTHING to do with the CPU.  Modern CPUs do not
communicate directly with a hard drive.  Some flavor of IDE, PATA, SATA or
SCSI interface handles data drives.

| Also, if I do buy the 4.7 GHz IBM Power6, I would like to undervolt and
| overclock it to the max possible. If I use the minimum voltage required
| for the CPU to operate, what is the highest clock frequency I can get?

| Just out of curiousity, is it physically-possible to build a CPU whose
| clock-rate is 600 PetaHz and whose Vcore is a picovolt?

The speed of light is an absolute limit to the operation of a single CPU.  A
petrahertz is 1/1,000,000,000,000,000 second in duration.  In that time NO
signal can travel further than 0.3 nanometer (less that the width of a
single transistor in a CPU.)

A picovolt is such an incredibly small voltage that likely someone combing
hair on the other side of the earth would change the local electrical field
by more than that.

As stated before, there is a minimum cut-off voltage for transistors - at
less than this voltage they do not "transist".  This voltage is about 0.7
volts for transistors used today; additional voltage is needed to make a
clear difference between the logic 0 state and the logic 1 state, additional
voltage is needed to raise the signal above the noise level.  Additional
voltage is needed to increased the switching speed.  The first transistors
were built on germanium rather than silicon; these germanium transistors
have a cutoff voltage of about 0.3 volts, but have many disadvantages,
including leakage, slower switching speeds, and higher power consumption.

The short answers are NO and NO.

Phil Weldon

| Ed M. wrote:
| >
| >>Phil Weldon wrote:
| >>
| >>
| >>>Overclocking is not going to damage an Intel x86 CPU before it becomes
| >>>obsolete (your Pentium 4 570J 3.8 GHz CPU already is.)  On the other
| >>>hand, raising the core voltage too high can immediately destroy an
Intel
| >>>x86 CPU. Limiting the core voltage to no more than +10% should give a
| >>>good safety margin; more than that is possible (there are many accounts
| >>>of successful overclocks with a higher core voltage increase, but there
| >>>are no guarantees.
| >>
| >>Is it possible to decrease the core voltage and at the same time,
increase
| >>the clock frequency?
| >>
| >
| >
| > In almost every case, NO.
| >
| >
| >>>There is no such thing as 'extreme undervolting.'  There is a cutoff
| >>>voltage at which a transistor will no longer operate.  Reducing the
core
| >>>voltage will not harm a CPU, but you will rapidly reach a voltage at
| >>>which the CPU will just not operate.  The CPU will not be damaged, nor
| >>>will any other part of the system.  Restoring the core voltage to
normal
| >>>will set everything aright.
| >>
| >>Okay. What if I use the minimum voltage necessary?
| >>
| >
| >
| > You can try it........ That processor is pretty much set to operate on
it's
| > default Vcore and lowering it will probably make it unstable even at
it's
| > default speed.
| >
| >
| >>>Read Phil's answer carefully, and also other posts in this newsgroup.
| >>>You will benefit from a better grasp of what overclocking entails.
| >>>Basically, just trading off voltage and heat safety margins for higher
| >>>clock speeds. Slowly and methodically is the way to go.
| >>
| >>Okay.
| >>
| >>My goal here is to use the minimum voltage required and at the same
time,
| >>the highest frequency possible. Is this possible?
| >
| >
| >     The best you could probably do is to find the highest possible freq
that
| > the processor will do at default Vcore, which will probably not be very
much
| > higher. What folks are trying to tell you is that the two processors you
| > mention are not the most efficient and not great OCers without voltage
| > INCREASES. They are very marginal even at default speeds and Vcore.
Those
| > processors were Intel's last go-round at using raw speed rather than
better
| > efficiency (operations per cycle) like the Core 2 Duos and Quads. An
E6600
| > at 2.4Ghz outperforms my EM64T 630 OC'd to 3.6Ghz by almost double in
any
| > benchmark. The Q6600 does even better, even in apps that are not
optimized
| > for it. Almost everyone in this group are very knowledgeable and are
really
| > trying to help you out. You just have to listen to what they are telling
| > you........Regards
| >
| >
| >
| > Ed
| >
| >
|
| It's true that CPUs with lower clock rates can be more efficient than
| CPUs with higher clock rates if the former uses more bits-per-cycle than
| the latter.
|
| As I've recently found, the 4.7 GHz IBM Power6 holds the current record
| for highest frequency CPU:
|
| http://www.spec.org/cpu2006/results/res2007q3/cpu2006-20070708-01383.html
|
| How much does the 4.7 GHz IBM Power6 cost?
|
| Is the 4.7 GHz IBM Power6 available today? If so, where can I find it? I
| live in Southern California in Diamond Bar. Any store close to me that
| carries this processor?
|
| What is the least expensive motherboard that is fully-compatible with
| this CPU?
|
| What is the least expensive HDD that is fully-compatible with this CPU?
|
| Also, if I do buy the 4.7 GHz IBM Power6, I would like to undervolt and
| overclock it to the max possible. If I use the minimum voltage required
| for the CPU to operate, what is the highest clock frequency I can get?
|
| I believe Power6 uses more instructions-per-cycle than most Intel or AMD
|  processors. So it is a combination of higher-clock-rate and more
| bits-per-cycle.
|
| Just out of curiousity, is it physically-possible to build a CPU whose
| clock-rate is 600 PetaHz and whose Vcore is a picovolt?



Re: Overclock but undervolt -- safe? How much?



Quoted text here. Click to load it

C'mon guys, if this isn't a troll I'll eat my C2D. Everyone's feeding him
lots.

(For some reason I don't see his posts, must be in my bit-bin from a
previous incarnation).
--
TTFN

Shaun.



Re: Overclock but undervolt -- safe? How much?



~misfit~ wrote:
Quoted text here. Click to load it

Save your digestion. Radium strings together a random
selection of science terms,the sillier the better and
then sits back to watch.
He is doing it on a wide selection of newsgroups.

Re: Overclock but undervolt -- safe? How much?



Somewhere on the interweb "Sjouke Burry" typed:
Quoted text here. Click to load it

I guess if one doesn't have a life, it keeps you occupied.

Thanks for the confirm (and saving me indigestion). <g>
--
TTFN

Shaun.



Re: Please answer me -- overclock but undervolt -- safe? How much?



Asked and answered the first time you posted.  And, by the way, you don't
set the terms of a conversation here.  A newsgroup is a group effort.  Most
participants in this group are very helpful and polite.  In fact, yours is
the most objectionable post I've seen here in years.

Phil Weldon

| Hi:
|
| Please answer my questions below. Please don't neglect this message.
|
| Is it possible to significantly decrease the heat generated by a
| "Pentium 4 570J/670 3.8 GHz processor" by using extreme undervolting
| without underclocking? How far can undervolting go before it causes
| trouble? What problems can result from undervolting? Can it damage the
| CPU or other parts of the PC?
|
| Also, can these processors be overclocked to 4 GHz while still being
| undervolted to the max tolerable? Does overclocking decrease the extent
| to which undervolting can be safely performed?
|
| In addition, what is the max that these CPUs can be overclocked without
| damaging them or decreasing their lifetimes?
|
| No offense but please respond with reasonable answers & keep out the
| jokes, off-topic nonsense, taunts, insults, and trivializations. I am
| really interested in this.
|
| If I am not answered properly I might go insane and do something
| everyone -- including me -- will eternally regret.
|
|
| Thanks,
|
| Radium



Re: Please answer me -- overclock but undervolt -- safe? How much?



Quoted text here. Click to load it

It's amusing to see that sort of complaint from someone top-posting.

--
Nate Edel                               http://www.cubiclehermit.com /

"What's the use of yearning for Elysian Fields when you know you can't get
'em, and would only let 'em out on building leases if you had 'em?" (WSG)

Re: Please answer me -- overclock but undervolt -- safe? How much?



'Nate Edel' wrote:
| It's amusing to see that sort of complaint from someone top-posting.
_____

I didn't feel like making the effort to trim the initial post, so consider
that my post came to me in a dream from Unknown Kadath

Phil Weldon


| > Asked and answered the first time you posted.  And, by the way, you
don't
| > set the terms of a conversation here.  A newsgroup is a group effort.
Most
| > participants in this group are very helpful and polite.  In fact, yours
is
| > the most objectionable post I've seen here in years.
|
| It's amusing to see that sort of complaint from someone top-posting.
|
| --
| Nate Edel                               http://www.cubiclehermit.com /
|
| "What's the use of yearning for Elysian Fields when you know you can't get
| 'em, and would only let 'em out on building leases if you had 'em?" (WSG)



Re: Please answer me -- overclock but undervolt -- safe? How much?



Top posting always seemed more logical to me.
Why should I have to scroll down through all the stuff I already read in the
last 20 posts.

Amir



Quoted text here. Click to load it




Re: Please answer me -- overclock but undervolt -- safe? How much?



Amir Facade wrote:
Quoted text here. Click to load it
Translation:Amir cares nothing about the responses
of other people and does not care about mixing up
said responses in a random order.
Does that describe a sociopath???
Or a google groups nitwit?????

Re: Please answer me -- overclock but undervolt -- safe? How much?



Somewhere on the interweb "Sjouke Burry" typed:
Quoted text here. Click to load it

I prefer the term "Googletard". :-)
--
TTFN

Shaun.



Re: Please answer me -- overclock but undervolt -- safe? How much?



wrote:
 :  Somewhere on the interweb "Sjouke Burry" typed:
 : > Amir Facade wrote:
 : > > Top posting always seemed more logical to me.
 : > > Why should I have to scroll down through all the stuff I already
 : > > read in the last 20 posts.
 : > >
 : > > Amir
 : > Translation:Amir cares nothing about the responses
 : > of other people and does not care about mixing up
 : > said responses in a random order.
 : > Does that describe a sociopath???
 : > Or a google groups nitwit?????
 :
 :  I prefer the term "Googletard". :-)

I've become an advocate of top "posting" in _emails_ with someone I
have to reply.  My wife read through some of the email exchanges that
invariably descended into nastiness.  She suggested that even though
I'm reponding appropriately the recipient is taking it as combative to
have the contradictions right there, inline, beneath the quoted bits
from the email eliciting the reply.  The combativeness is pretty much
gone from those email exchanges now.

email, not on a netnews froup...

Re: Please answer me -- overclock but undervolt -- safe? How much?



If you drop the voltage to the CPU you may fry it because it will try to
draw more current in order to run itself, and, then, toast.  NOT a good
idea.

--
---------------------
DaveW

---------------------
Quoted text here. Click to load it



Re: Please answer me -- overclock but undervolt -- safe? How much?




Quoted text here. Click to load it
    Rubbish again David. Todays processors will just become unstable or not
operate at lower voltages. They will NOT fry. Going too high with the
voltages is another matter.


Ed



Re: Please answer me -- overclock but undervolt -- safe? How much?



nothing@bot.org says...
Quoted text here. Click to load it
BS.

C2D 6420 stable at 300 x 8 and 1.06 volts. Cooler than ever.

--
http://ardennes.free.fr

Re: Please answer me -- overclock but undervolt -- safe? How much?



'DaveW' wrote:
| If you drop the voltage to the CPU you may fry it because it will try to
| draw more current in order to run itself, and, then, toast.  NOT a good
| idea.
_____

The lower the voltage, the LOWER the power drawn, the LOWER the heat
dissipated and the LOWER the stress on the CPU.  Not to mention that once
you drop the CPU core voltage too low, it just will not operate until you
restore the core voltage to something nearer the Intel specified voltage.

It would be a GOOD idea for you to do some investigation BEFORE posting
another of your howlers.  On the other hand, if your posts are meant to be
funny, keep up the good work!

Phil Weldon

| If you drop the voltage to the CPU you may fry it because it will try to
| draw more current in order to run itself, and, then, toast.  NOT a good
| idea.
|
| --
| ---------------------
| DaveW
|
| ---------------------



Site Timeline