Lowering the voltage of the CPU

[Stegozor]

Hi everyone,

The latest BOINC UK newsletter
(http://www.boincuk.com/newsLetter/bunc-03.pdf 212 K) has an interesting
article in the 7th page. It's about keeping the CPU cool, and there's a
section about lowering the CPU voltage in the BIOS. The author says he
managed to lower the CPU temperature without any loss of performance on
his computer. As lowering the temperature makes your CPU last longer,
I'm interested in doing the same, but I wonder if this is feasible on my
hardware (Asus A7V333 motherboard, AMD Athlon XP 2000+ running at 1.1
Ghz - voluntary BOIS setting -, the whole thing under Ubuntu GNU/Linux),
and also if it's reasonable to do so.

Any advice would be more than welcome. TIA.

 

[Paul]

Hi everyone,

The latest BOINC UK newsletter
(http://www.boincuk.com/newsLetter/bunc-03.pdf 212 K) has an interesting
article in the 7th page. It's about keeping the CPU cool, and there's a
section about lowering the CPU voltage in the BIOS. The author says he
managed to lower the CPU temperature without any loss of performance on
his computer. As lowering the temperature makes your CPU last longer,
I'm interested in doing the same, but I wonder if this is feasible on my
hardware (Asus A7V333 motherboard, AMD Athlon XP 2000+ running at 1.1
Ghz - voluntary BOIS setting -, the whole thing under Ubuntu GNU/Linux),
and also if it's reasonable to do so.

Any advice would be more than welcome. TIA.

Asus BIOS, in general, do not allow undervolting. If AMD specifies
1.65 +/- 0.05 for the voltage for a certain AthlonXP, then the BIOS
may have a 1.60 setting and that is as low as you can go. There are
ways to fix that, from a hardware perspective. But they are not
practical ways. I've had only one poster who went to the trouble of
setting his voltage on an Asus motherboard, with a hardware mod.

AMD designed your processor, to operate properly and have a long
life with a specified voltage and temperature operating region. While
you could be right, that dropping the voltage gives a longer life,
there are plenty of other things that can die and fry your system
(failing PSU, bad motherboard capacitors etc). If the processor
doesn't die, something else will.

There are other brands of motherboards that do allow undervolting
the processor.

And if the processor dies, you have an opportunity to upgrade, which
is not all bad. Think of running a quad core, and giving three cores
to Boinc, while you use the other one.

Paul

 

[kony]

Asus BIOS, in general, do not allow undervolting. If AMD specifies
1.65 +/- 0.05 for the voltage for a certain AthlonXP, then the BIOS
may have a 1.60 setting and that is as low as you can go. There are
ways to fix that, from a hardware perspective. But they are not
practical ways. I've had only one poster who went to the trouble of
setting his voltage on an Asus motherboard, with a hardware mod.

I've set the voltage on many Asus boards with hardware mods,
actually too many to remember but they include A7V600,
A7N8X, and A7N266 (Not sure about this model #, was a mATX
nForce1 board). Even more earlier era boards but these are
roughly contemporary.

The OP's board on the other hand, A7V333, has a better
flexibility than average. IIRC, it has bios voltage
settings but it also has a row of headers right above the
CPU socket which manipulate the VID pin signals... so with
these jumpers you can override the supposed CPU default
voltage, and even then adjust it further with the bios
settings. This allows the full range of possible voltages,
from somewhere around 1.3V (might even be down closer to 1.0
or 1.1V, not sure), up to at least 1.85V if not higher.

I suppose it's possible that if OP has an "-X" version, like
A7V333-X, or an OEM version, it might leave the voltage
jumpers off the board entirely (though possible still viable
by soldering on the pins or just jumper wires).

Correction, I just found a note in a docs collection for the
A7V333 and it states:

===========================

The V/Core can be upped to 2.24V by enabling the "over1"
jumper found beside the CPU socket.

===========================

Perhaps most useful is to just use this diagram, which is a
separate jumper bank from aforementioned "over1" jumper.

http://69.36.189.159/usr_1034/A7V333_Jumpers1.gif

AMD designed your processor, to operate properly and have a long
life with a specified voltage and temperature operating region. While
you could be right, that dropping the voltage gives a longer life,
there are plenty of other things that can die and fry your system
(failing PSU, bad motherboard capacitors etc). If the processor
doesn't die, something else will.

There are other brands of motherboards that do allow undervolting
the processor.

This particular Asus board is a wierd blend of both bios
settings and a very large assortment of jumpers, it might be
an all time record for # of jumpers for Asus. They make it
particularly easy to over or underclock without having to
reset bios for FSB or voltage... it's just a pity it has the
Via KT333 chipset which was it's weakest link, IMO.

 

[kony]

Hi everyone,

The latest BOINC UK newsletter
(http://www.boincuk.com/newsLetter/bunc-03.pdf 212 K) has an interesting
article in the 7th page. It's about keeping the CPU cool, and there's a
section about lowering the CPU voltage in the BIOS.

First, are you having trouble keeping it cool? What is the
purpose for this change, the realized benefit you expect
from lowering the voltage? I'm not claiming there aren't
any potential benefits, but a bit more detail about your
specific goal might help.

The most thorough way of doing it is to plot a graph, of
voltage vs. CPU speed. Through trial and error you'd find
the max speed the CPU can run stabily (Prime95's Torture
Test, Large In-Place FFTs setting ran for at least several
hours or terminated for readjustment of settings immediately
if an error is found) at each voltage setting available.

Having this data you can decide what tradeoff to make. Keep
in mind that your board's KT333 chipset doesn't have an AGP
(and PCI) lock on the FSB, so you need to stay fairly close
to the 100/133/166 FSB speeds your board supports (though
166 is unofficial, at 166 and especially going towards
175MHz FSB you may find your AGP video cards start acting
flaky). Point being, it would be somewhat arbitrary to only
think about keeping the CPU at it's present speed and
lowering the voltage as low as it can go while remaining
stable. It might even be possible to raise the CPU speed,
overclock slightly, while still lowering the voltage and
even have lower heat as a result.

On the other hand, for even more heat reduction it might
help to decide on your target voltage, and THEN find the
peak CPU speed that can run stable at that voltage (be sure
to test the final config overnight running Prime95 Test as
mentioned above, before ever booting windows/other OS).

The author says he
managed to lower the CPU temperature without any loss of performance on
his computer.

Sure, if you keep the clock speeds and multiplier the same
and only lower voltage, that will retain performance- so
long as it stays stable which is the crux of the problem,
that if ALL of your model of CPU stayed stable at lower
voltage than AMD had spec'd, they might've just spec'd them
all for lower voltage instead of that they did use. Each
specimen of CPU may respond a little differently, and while
AMD tried to make it easier for users or motherboard
manufacturers by keeping same voltage for multiple speeds in
same CPU family, each may have a little bit different
voltage:frequency curve, also depending on how cool your
heatsink/case/ambient temp, keep it as well.

As lowering the temperature makes your CPU last longer,
I'm interested in doing the same,

Unless your CPU is constantly overheating (say past 65-70C
which is even then more of a stability concern than a
lifespan concern), it's lifespan should be fine without any
attempt to lower the voltage-> heat.

but I wonder if this is feasible on my
hardware (Asus A7V333 motherboard, AMD Athlon XP 2000+ running at 1.1
Ghz - voluntary BOIS setting -, the whole thing under Ubuntu GNU/Linux),
and also if it's reasonable to do so.

Yes, your board is one of the easiest ever made to undervolt
Athlons, at least if you don't mind having the case open and
manipulating jumpers. Use the voltage jumpers as I'd
mentioned in reply to Paul's post, and see this diagram.

http://69.36.189.159/usr_1034/A7V333_Jumpers1.gif

It's been awhile since I fooled around with one of those
A7V333 boards, but I vaguely recall that I did test the
pre-1.5V jumper settings, but on that board with it's
through-hole heatsink mounting system, I had an Alpha
Pal8045 heatsink which was pretty good, it cooled
sufficiently at very low noise so I had no need to keep it
undervolted, think it ended up runnning a pre-release
Thornton @ 1.65V/2.2Ghz/172MHz FSB on it but it was too
picky about memory, wouldn't take several pairs of 512MB
DIMMs I tried at that (172MHz) synchronous memory bus speed
and I reserved all my better memory for nForce2 board uses
since they could exploit memory so much more at the time.

Anyway, what voltage you can hit at 1.1GHz may depend on
which (CPU) core you have, Palomino might still need
1.3-1.4V or so but T-Bred -A, especially -B or Thorton core
might get down pretty close to 1.1V... but frankly unless
you're trying to passively cool the CPU, I don't see the
need to underclock and undervolt it that much unless it's
operating in a quite hostile (high temp) ambient
environment.

 

[kony]

On the other hand, for even more heat reduction it might
help to decide on your target voltage, and THEN find the
peak CPU speed that can run stable at that voltage (be sure
to test the final config overnight running Prime95 Test as
mentioned above, before ever booting windows/other OS).

Hmm. I suppose you might want to boot the OS, since that's
how you'd run Prime95.

 

[meow2222]

Hi everyone,

The latest BOINC UK newsletter
(http://www.boincuk.com/newsLetter/bunc-03.pdf 212 K) has an interesting
article in the 7th page. It's about keeping the CPU cool, and there's a
section about lowering the CPU voltage in the BIOS. The author says he
managed to lower the CPU temperature without any loss of performance on
his computer. As lowering the temperature makes your CPU last longer,
I'm interested in doing the same, but I wonder if this is feasible on my
hardware (Asus A7V333 motherboard, AMD Athlon XP 2000+ running at 1.1
Ghz - voluntary BOIS setting -, the whole thing under Ubuntu GNU/Linux),
and also if it's reasonable to do so.

Any advice would be more than welcome. TIA.

is it the best thing you can do with your time?


NT

 

[paulmd]

Hi everyone,

The latest BOINC UK newsletter
(http://www.boincuk.com/newsLetter/bunc-03.pdf 212 K) has an interesting
article in the 7th page. It's about keeping the CPU cool, and there's a
section about lowering the CPU voltage in the BIOS. The author says he
managed to lower the CPU temperature without any loss of performance on
his computer. As lowering the temperature makes your CPU last longer,
I'm interested in doing the same, but I wonder if this is feasible on my
hardware (Asus A7V333 motherboard, AMD Athlon XP 2000+ running at 1.1
Ghz - voluntary BOIS setting -, the whole thing under Ubuntu GNU/Linux),
and also if it's reasonable to do so.

Any advice would be more than welcome. TIA.

If they don't overheat, or get overclocked, processors are near
immortal. I routinely see working 20 year old computers. Don't worry
about it.

 

[Stegozor]

First, are you having trouble keeping it cool? What is the
purpose for this change, the realized benefit you expect
from lowering the voltage? I'm not claiming there aren't
any potential benefits, but a bit more detail about your
specific goal might help.

Frankly, it depends. Two or three years ago, it was running at 1.667 Ghz
(the default setting (previous frequencies I mentioned from memory were
inaccurate). I was already crunching for seti@home classic. At some
point during the summer, the computer started to reboot with "cpu
temperature too high" vocal POST message. Then, I lowered its frequency
at 1.250 Ghz (rather than buying a new fan, laziness) and things went
well for quite a while. This summer (when the weather was hot, around
32°C), my lovely computer began to reboot again, with the same message.
So, since I read the BOINC UK newsletter, I wonder if lowering the CPU
voltage could prevent such a thing from happening again while
maintaining global performance. Now, there almost no problem, but I
think that next summer, things might go wrong again unless I find a
decent solution.

I'm not looking for ultimate performance, all I want is to continue to
calculate happily for BOINC, to ensure my CPU will last long, and avoid
such shut downs. Seems that I should definitely get a brand new fan...

Also, another question: in the BIOS, the Over Shut Down Setting (that
reboots the computer if CPU temperature is too hot) is set on Auto. When
I switch it on Manual, the default temp is 90°C. I hope Auto is not
waiting for 90°C be reached, as this seems pretty high to me.


Some additional information:

According to the BIOS' hardware monitor, just after a reboot following
normal use:

Motherboard temperature: 34°C/91°F
CPU temperature: 73.5°C/164.5°F
CPU Fan: around 2848 rpm
Power Fan: around 3479 rpm
Vcore Voltage: 1.79 V
+3.3V Voltage: 3.13~3.15 V
+5 Voltage: 4.75~4.80 V
+12 Voltage: 12.9 V

When I stop BOINC, it will turn to:
Motherboard temperature: 32°C/89.5°F
CPU temperature: 64.5°C/147.5°F
The other values remain unchanged.

http://69.36.189.159/usr_1034/A7V333_Jumpers1.gif

Nice diagram, thanks. I'll have a closer look at the details ASAP.
Thanks a lot.

 

[meow2222]

Frankly, it depends. Two or three years ago, it was running at 1.667 Ghz
(the default setting (previous frequencies I mentioned from memory were
inaccurate). I was already crunching for seti@home classic. At some
point during the summer, the computer started to reboot with "cpu
temperature too high" vocal POST message. Then, I lowered its frequency
at 1.250 Ghz (rather than buying a new fan, laziness) and things went
well for quite a while. This summer (when the weather was hot, around
32°C), my lovely computer began to reboot again, with the same message.

such shut downs. Seems that I should definitely get a brand new fan...

I do believe youve cracked it


NT

 

[kony]

Frankly, it depends. Two or three years ago, it was running at 1.667 Ghz
(the default setting (previous frequencies I mentioned from memory were
inaccurate). I was already crunching for seti@home classic.

Is this a primary use system or just crunching SETI with no
other purpose? I ask because it's possible you'd have more
efficient results if you overclock and only run the system
when it's cooler (like at night or morning), then shut off
the system other times. Higher CPU clock speed reduces the
total power consumption a little %, while turning system off
entirely is a great reduction, barely any power used for
5VSB.

At some
point during the summer, the computer started to reboot with "cpu
temperature too high" vocal POST message.

We'd need more specifics about this... it's possible to set
that shutoff temp too low, and it's also possible your board
is reading the CPU temp higher than it actually is. Point
being, your system might be shutting off when the CPU wasn't
actually too hot yet. I'd find out the shutdown temp,
whether your board reports CPU temp a few degrees high (I
think it may be about 5C higher than actual but I might be
confusing this with a different Asus board) and whether
anything else is getting too hot- it might just be that you
need more chassis airflow, that underclocking the CPU might
resolve the shutdown temp issue but did nothing about other
parts similarly running hotter when your ambient temp is
higher.

Then, I lowered its frequency
at 1.250 Ghz (rather than buying a new fan, laziness) and things went
well for quite a while. This summer (when the weather was hot, around
32°C), my lovely computer began to reboot again, with the same message.
So, since I read the BOINC UK newsletter, I wonder if lowering the CPU
voltage could prevent such a thing from happening again while
maintaining global performance. Now, there almost no problem, but I
think that next summer, things might go wrong again unless I find a
decent solution.

Well you're not maintaining performance if running at
1.25GHz, but IF you find your CPU is running hot (maybe
above 65C during SETI use, regardless of what the bios or
software shutdown setting is), and your chassis has good
airflow, and you don't think a CPU heatsink upgrade would
help enough, then I'd agree that undervolting might be a
solution, unless you instead chose to do as mentioned above,
raise the clockspeed and just turn it off during the hotter
parts of the day.

I'm not looking for ultimate performance, all I want is to continue to
calculate happily for BOINC, to ensure my CPU will last long, and avoid
such shut downs. Seems that I should definitely get a brand new fan...

Ok, but ultimate performance is essentially how to do same
calculations in less time, so it's not needing to run all
the time which saves energy too. Your call.

Also, another question: in the BIOS, the Over Shut Down Setting (that
reboots the computer if CPU temperature is too hot) is set on Auto. When
I switch it on Manual, the default temp is 90°C. I hope Auto is not
waiting for 90°C be reached, as this seems pretty high to me.

Why not check the temp?
I'd set it to 70C or higher, but 90C does seem too high.

Some additional information:

According to the BIOS' hardware monitor, just after a reboot following
normal use:

Motherboard temperature: 34°C/91°F
CPU temperature: 73.5°C/164.5°F
CPU Fan: around 2848 rpm
Power Fan: around 3479 rpm
Vcore Voltage: 1.79 V
+3.3V Voltage: 3.13~3.15 V
+5 Voltage: 4.75~4.80 V
+12 Voltage: 12.9 V

When I stop BOINC, it will turn to:
Motherboard temperature: 32°C/89.5°F
CPU temperature: 64.5°C/147.5°F
The other values remain unchanged.

That's at stock speed or at 1.1 or 1.25 GHz?
What's your room ambient temp and what heatsink are you
using? Just trying to get an idea of where your weaker
links are in cooling...

Based on the 1.79V reported above (which is likely meaning
the board is overvoltage a bit from the potential, default
1.75V of the CPU), it does seem you have a Palomino core
Athlon, unless you had manually set it to a higher voltage
than stock. I'd try 1.5V, that might be about right for
1.25GHz or a little higher and if you keep the frequency the
same it should result in a little less than 75% of your
current CPU heat production.

You might also consider removing the heatsink and applying a
fresh thin coat of thermal compound if you hadn't recently,
especially if the original was generic silicon grease which
tends to dry out over time, especially at those temps on an
open flipchip core like an Athlon XP has.

Nice diagram, thanks. I'll have a closer look at the details ASAP.
Thanks a lot.

I suggest leaving the bios settings for voltage at default,
and entirely using these jumpers so if the board fails to
post, you have no need to clear CMOS, just increase the
voltage by 0.05V increments until you hit your target temp
during stable Prime95 (or SETI of course) operation.

 

[Stegozor]

Is this a primary use system or just crunching SETI with no
other purpose? (...)

BOINC/SETI runs automagically each time I log in and while I normally
use my computer. I do nothing exciting, web browsing with SeaMonkey,
some OOo, a few basic games from time to time, and listening to music.
In the evening I turn the computer off, I mean, I don't leave it on just
for BOINC.

We'd need more specifics about this... it's possible to set
that shutoff temp too low, and it's also possible your board
is reading the CPU temp higher than it actually is. Point
being, your system might be shutting off when the CPU wasn't
actually too hot yet. I'd find out the shutdown temp,
whether your board reports CPU temp a few degrees high (I
think it may be about 5C higher than actual but I might be
confusing this with a different Asus board) and whether
anything else is getting too hot- it might just be that you
need more chassis airflow, that underclocking the CPU might
resolve the shutdown temp issue but did nothing about other
parts similarly running hotter when your ambient temp is
higher.

Except an ATI Radeon 9200 video card, I don't see what would cause such
a rise of temperature. I also hope that my motherboard isn't reporting a
flawed temp since I don't know how to get correct values otherwise.

Well you're not maintaining performance if running at
1.25GHz, but IF you find your CPU is running hot (maybe
above 65C during SETI use, regardless of what the bios or
software shutdown setting is), and your chassis has good
airflow, and you don't think a CPU heatsink upgrade would
help enough, then I'd agree that undervolting might be a
solution, unless you instead chose to do as mentioned above,
raise the clockspeed and just turn it off during the hotter
parts of the day.

OK, so undervolting seems to be the last possibility to consider. I
shoud get a new fan I shoud get a new fan I shoud get a new fan I shoud
get a new fan...

That's at stock speed or at 1.1 or 1.25 GHz?

At 1.25 Ghz, the usual value since the first "temp too high" messages.
(1.25 & 1.667 are the only choices available within the BIOS).

What's your room ambient temp and what heatsink are you

The ambient temp in the room right now is probably around 20~25°C. I
just feel good

using? Just trying to get an idea of where your weaker
links are in cooling...

The heatsink and the CPU fan seem to be basic nonames. We're talking
about a computer built with second hand components. Money problems, you
know. Otherwise, I'd gladly but a multicore 64 bit AMD...

Based on the 1.79V reported above (which is likely meaning
the board is overvoltage a bit from the potential, default
1.75V of the CPU), it does seem you have a Palomino core
Athlon, unless you had manually set it to a higher voltage
than stock. I'd try 1.5V, that might be about right for

Nope, I didn't dare to play with the voltage yet The lowest voltage
allowed by the BIOS is 1.75 V, and even when I switch on manual setting
and choose 1.75 (normally I'm on Auto), the CPU Vcore voltage oscillates
around 1.79 V.

1.25GHz or a little higher and if you keep the frequency the
same it should result in a little less than 75% of your
current CPU heat production.

You might also consider removing the heatsink and applying a
fresh thin coat of thermal compound (...)

Recently I cleaned the heatsink and the CPU, GPU and power supply fans
(all the dust you can remove, it's incredible!) but surprisingly with no
noticeable change. Anyway, the thermal compound is a nice idea.