Re: New Theory: AMD's CPU Throttling damaged the power supply.
Bookmark this page:
 Yahoo!
 Google
 Windows Live
 del.icio.us
 digg
 Netscape
|
|
|||||||||||||
|
Posted by Archimedes' Lever on August 4, 2008, 4:33 am
Please log in for more thread options >> > > Hello,
>>
>> > > I have a new theory how the power supply got damaged:
>>
>> > > AMD's X2 3800+ dual core processor has CPU Throttling.
>>
>> > > I downloaded a special AMD Software tool called AMD Power Monitor or
>> > > something like that. >>
>> > > Which throttles back the CPU when it's idling... then when it needs to do
>> > > work the cpu is throttled up. >>
>> > > So on idle the processor will work at 1.0 volts... and then when it's busy
>> > > it will work on 2.0 volts... maybe even voltages between it. >>
>> > > These voltages fluctuations occur very rapidly.
>>
>> > > Think of it like this:
>>
>> > > You turn on the water... you turn off the water... then you turn it on
>> > > again... then you turn it off again. >>
>> > > If you do this at the right pace... the water in the tubes will start to
>> > > pound on the inner tubing... and sooner or later something is going to >> > > break. >>
>> > > This is probably what is happening inside the electronics.
>>
>> > > It will work for a while... but after one or two years of heavy usage...
>> > > kiss it goodbye ! >>
>> > > Fluctuations in temperature seem to also cause problems for gpu's in
>> > > laptops... >>
>> > > Try doing that with your teeth fillings... ouch ;)
>>
>> > > Crappy hardware nowadays, me very sorry I bought this crap.
>>
>> > > But it is powerfull and with some nice Skybuck experience and knowledge it
>> > > might work after all. >>
>> > > So I gotta find a new power supply... and leave the processor at always
2.0
>> > > volts to prevent any damage or temperature fluctuations.
>>
>> > > Don't know what happens to gpu though... though the 7900 gtx seems to be
on
>> > > able to take quiet a beating... don't know about the components around it >> > > the motherboard though ;)
>>
>> > > You've been warned people =D
>>
>> > > Bye,
>> > > Skybuck. >>
>> > The clock rate is throttled, not the supply voltage.
>>
>> If you throttle the clock rate the processor voltage can be as well. >> Reducing the clock rate reduces power less than linearly (constant >> leakage, which is significant). Reducing the voltage reduces the >> power by the square. >> >> -- >> Keith >
>I don't see this being done on desktop CPUs. That is ridiculous. A HUGE number of motherboards not only allow memory timing adjustments. Most, if not all of those also have provisions for reducing or raising the voltage setpoint of specific rails.. | |||||||||||||
|
Posted by on August 5, 2008, 1:21 am
Please log in for more thread options > On Sun, 3 Aug 2008 19:45:30 -0700 (PDT), m...@sushi.com wrote:
> >> In article <bce1d2f3-1e58-430e-be16-22883e1b6cd7
> >> @r35g2000prm.googlegroups.com>, m...@sushi.com says... >
e:
> >> > > Hello,
>
> >> > > I have a new theory how the power supply got damaged:
>
> >> > > AMD's X2 3800+ dual core processor has CPU Throttling.
>
> >> > > I downloaded a special AMD Software tool called AMD Power Monitor =
or
> >> > > something like that.
>
> >> > > Which throttles back the CPU when it's idling... then when it need=
s to do
> >> > > work the cpu is throttled up.
>
> >> > > So on idle the processor will work at 1.0 volts... and then when i=
t's busy
> >> > > it will work on 2.0 volts... maybe even voltages between it.
>
> >> > > These voltages fluctuations occur very rapidly.
>
> >> > > Think of it like this:
>
> >> > > You turn on the water... you turn off the water... then you turn i=
t on
> >> > > again... then you turn it off again.
>
> >> > > If you do this at the right pace... the water in the tubes will st=
art to
> >> > > pound on the inner tubing... and sooner or later something is goin=
g to
> >> > > break.
>
> >> > > This is probably what is happening inside the electronics.
>
> >> > > It will work for a while... but after one or two years of heavy us=
age...
> >> > > kiss it goodbye !
>
> >> > > Fluctuations in temperature seem to also cause problems for gpu's =
in
> >> > > laptops...
>
> >> > > Try doing that with your teeth fillings... ouch ;)
>
> >> > > Crappy hardware nowadays, me very sorry I bought this crap.
>
> >> > > But it is powerfull and with some nice Skybuck experience and know=
ledge it
> >> > > might work after all.
>
> >> > > So I gotta find a new power supply... and leave the processor at a=
lways 2.0
> >> > > volts to prevent any damage or temperature fluctuations.
>
> >> > > Don't know what happens to gpu though... though the 7900 gtx seems=
to be
> >> > > able to take quiet a beating... don't know about the components ar=
ound it on
> >> > > the motherboard though ;)
>
> >> > > You've been warned people =3DD
>
> >> > > Bye,
> >> > > =A0 Skybuck. >
> >> > The clock rate is throttled, not the supply voltage.
>
> >> If you throttle the clock rate the processor voltage can be as well. =
=A0
> >> Reducing the clock rate reduces power less than linearly (constant
> >> leakage, which is significant). =A0Reducing the voltage reduces the > >> power by the square. >
> >> --
> >> Keith >
> >I don't see this being done on desktop CPUs.
>
s.
> That is ridiculous. > > =A0 A HUGE number of motherboards not only allow memory timing adjustment= > Most, if not all of those also have provisions for reducing or raising
> the voltage setpoint of specific rails.. The key here is changing voltage on the fly. Your argument is irrelevant. | |||||||||||||
|
Posted by Archimedes' Lever on August 5, 2008, 4:38 am
Please log in for more thread options On Mon, 4 Aug 2008 22:21:03 -0700 (PDT), miso@sushi.com wrote:
>wrote:
>> On Sun, 3 Aug 2008 19:45:30 -0700 (PDT), m...@sushi.com wrote:
>> >> In article <bce1d2f3-1e58-430e-be16-22883e1b6cd7
>> >> @r35g2000prm.googlegroups.com>, m...@sushi.com says... >>
>> >> > > Hello,
>>
>> >> > > I have a new theory how the power supply got damaged:
>>
>> >> > > AMD's X2 3800+ dual core processor has CPU Throttling.
>>
>> >> > > I downloaded a special AMD Software tool called AMD Power Monitor or
>> >> > > something like that. >>
>> >> > > Which throttles back the CPU when it's idling... then when it needs to
do
>> >> > > work the cpu is throttled up.
>>
>> >> > > So on idle the processor will work at 1.0 volts... and then when it's
busy
>> >> > > it will work on 2.0 volts... maybe even voltages between it.
>>
>> >> > > These voltages fluctuations occur very rapidly.
>>
>> >> > > Think of it like this:
>>
>> >> > > You turn on the water... you turn off the water... then you turn it on
>> >> > > again... then you turn it off again. >>
>> >> > > If you do this at the right pace... the water in the tubes will start
to
>> >> > > pound on the inner tubing... and sooner or later something is going to
>> >> > > break. >>
>> >> > > This is probably what is happening inside the electronics.
>>
>> >> > > It will work for a while... but after one or two years of heavy
usage...
>> >> > > kiss it goodbye !
>>
>> >> > > Fluctuations in temperature seem to also cause problems for gpu's in
>> >> > > laptops... >>
>> >> > > Try doing that with your teeth fillings... ouch ;)
>>
>> >> > > Crappy hardware nowadays, me very sorry I bought this crap.
>>
>> >> > > But it is powerfull and with some nice Skybuck experience and
knowledge it
>> >> > > might work after all.
>>
>> >> > > So I gotta find a new power supply... and leave the processor at
always 2.0
>> >> > > volts to prevent any damage or temperature fluctuations.
>>
>> >> > > Don't know what happens to gpu though... though the 7900 gtx seems to
be
>> >> > > able to take quiet a beating... don't know about the components around
it on
>> >> > > the motherboard though ;)
>>
>> >> > > You've been warned people =D
>>
>> >> > > Bye,
>> >> > > Skybuck. >>
>> >> > The clock rate is throttled, not the supply voltage.
>>
>> >> If you throttle the clock rate the processor voltage can be as well.
>> >> Reducing the clock rate reduces power less than linearly (constant >> >> leakage, which is significant). Reducing the voltage reduces the >> >> power by the square. >>
>> >> --
>> >> Keith >>
>> >I don't see this being done on desktop CPUs.
>>
>> That is ridiculous. >> >> A HUGE number of motherboards not only allow memory timing adjustments. >> Most, if not all of those also have provisions for reducing or raising >> the voltage setpoint of specific rails.. >
>The key here is changing voltage on the fly. Your argument is >irrelevant. You are irrelevant. "Dynamic, on-the-fly rails" and digital logic doesn't go together, dingledorf. | |||||||||||||
|
Posted by krw on August 5, 2008, 6:00 am
Please log in for more thread options OneBigLever@InfiniteSeries.Org says...
> On Mon, 4 Aug 2008 22:21:03 -0700 (PDT), miso@sushi.com wrote:
>=20 > >wrote:
> >> On Sun, 3 Aug 2008 19:45:30 -0700 (PDT), m...@sushi.com wrote:
rote:
> >> >> In article <bce1d2f3-1e58-430e-be16-22883e1b6cd7 > >> >> @r35g2000prm.googlegroups.com>, m...@sushi.com says... > >> > >> >> > > Hello,
or or
> >> > >> >> > > I have a new theory how the power supply got damaged: > >> > >> >> > > AMD's X2 3800+ dual core processor has CPU Throttling. > >> > >> >> > > I downloaded a special AMD Software tool called AMD Power Monit= > >> >> > > something like that.
eeds to do
> >> > >> >> > > Which throttles back the CPU when it's idling... then when it n= > >> >> > > work the cpu is throttled up.
n it's busy
> >> > >> >> > > So on idle the processor will work at 1.0 volts... and then whe= > >> >> > > it will work on 2.0 volts... maybe even voltages between it.
n it on
> >> > >> >> > > These voltages fluctuations occur very rapidly. > >> > >> >> > > Think of it like this: > >> > >> >> > > You turn on the water... you turn off the water... then you tur= > >> >> > > again... then you turn it off again.
start to
> >> > >> >> > > If you do this at the right pace... the water in the tubes will= > >> >> > > pound on the inner tubing... and sooner or later something is g=
oing to
> >> >> > > break.
usage...
> >> > >> >> > > This is probably what is happening inside the electronics. > >> > >> >> > > It will work for a while... but after one or two years of heavy= > >> >> > > kiss it goodbye !
's in
> >> > >> >> > > Fluctuations in temperature seem to also cause problems for gpu= > >> >> > > laptops...
nowledge it
> >> > >> >> > > Try doing that with your teeth fillings... ouch ;) > >> > >> >> > > Crappy hardware nowadays, me very sorry I bought this crap. > >> > >> >> > > But it is powerfull and with some nice Skybuck experience and k= > >> >> > > might work after all.
t always 2.0
> >> > >> >> > > So I gotta find a new power supply... and leave the processor a= > >> >> > > volts to prevent any damage or temperature fluctuations.
ems to be
> >> > >> >> > > Don't know what happens to gpu though... though the 7900 gtx se= > >> >> > > able to take quiet a beating... don't know about the components=
around it on
> >> >> > > the motherboard though ;)
. =A0
> >> > >> >> > > You've been warned people =3DD > >> > >> >> > > Bye, > >> >> > > =A0 Skybuck. > >> > >> >> > The clock rate is throttled, not the supply voltage. > >> > >> >> If you throttle the clock rate the processor voltage can be as well= > >> >> Reducing the clock rate reduces power less than linearly (constant
ents.
> >> >> leakage, which is significant). =A0Reducing the voltage reduces the > >> >> power by the square. > >> > >> >> -- > >> >> Keith > >> > >> >I don't see this being done on desktop CPUs. > >> > >> That is ridiculous. > >> > >> =A0 A HUGE number of motherboards not only allow memory timing adjustm= > >> Most, if not all of those also have provisions for reducing or raising
> >> the voltage setpoint of specific rails.. > >
> >The key here is changing voltage on the fly. Your argument is > >irrelevant. >=20
>=20 > You are irrelevant. "Dynamic, on-the-fly rails" and digital logic > doesn't go together, dingledorf. Wrong again, AlwaysWrong. How *do* you do it, Dimbulb? Modulating=20 the power rails is a well known method for reducing power=20 consumption in laptops. ...unless you don't consider=20 microprocessors to be "digital logic". --=20 Keith | |||||||||||||
|
Posted by Archimedes' Lever on August 5, 2008, 8:35 pm
Please log in for more thread options
>
>Wrong again, AlwaysWrong. How *do* you do it, Dimbulb? Modulating >the power rails is a well known method for reducing power >consumption in laptops. ...unless you don't consider >microprocessors to be "digital logic". > >-- >Keith Sleep with a reduced rail voltage is quite different from full bore operation with sliding rails. So no. When the mobo throttles voltages, it also throttles the cloock, and no, that is NOT an example of a varying rail on an active digital pathway. | |||||||||||||
>> @r35g2000prm.googlegroups.com>, m...@sushi.com says...
>>
>>
>>