E6600 and D975XBX2

And does somebody nows something about this one?

The other thing is, when I look with CPU-Z is my multiplier switching
himself from 9x to 6x and back...
 >> Stay informed about: E6600 and D975XBX2 

'Tom' wrote:
| And does somebody nows something about this one?
|
| The other thing is, when I look with CPU-Z is my multiplier switching
| himself from 9x to 6x and back...
_____

My answer applies to the Intel E6600. What you see with CPU-Z is 'thermal
throttling'. The CPU, in a completely on-CPU-chip action that REDUCES the
CPU multiplier to lower heat production if the CPU temperature becomes too
high. You can disable 'thermal throttle' (automatic CPU multiplier
reduction [based on CPU temperature], but you can NOT directly change the
CPU multiplier. You can NOT INCREASE the CPU multiplier.

Phil Weldon

<tom DeleteThis @japtuning.ch> wrote in message
news:1172776523.511628.272630@t69g2000cwt.googlegroups.com...
| And does somebody nows something about this one?
|
| The other thing is, when I look with CPU-Z is my multiplier switching
| himself from 9x to 6x and back...
|
 >> Stay informed about: E6600 and D975XBX2 

Phil Weldon wrote:
> 'Tom' wrote:
> | And does somebody nows something about this one?
> |
> | The other thing is, when I look with CPU-Z is my multiplier switching
> | himself from 9x to 6x and back...
> _____
>
> My answer applies to the Intel E6600. What you see with CPU-Z is 'thermal
> throttling'. The CPU, in a completely on-CPU-chip action that REDUCES the
> CPU multiplier to lower heat production if the CPU temperature becomes too
> high. You can disable 'thermal throttle' (automatic CPU multiplier
> reduction [based on CPU temperature], but you can NOT directly change the
> CPU multiplier. You can NOT INCREASE the CPU multiplier.
>
> Phil Weldon

It could be EIST.

http://www.intel.com/cd/channel/reseller/asmo-na/eng/products/box_proc...ors/des

I haven't been able to find any info on how EIST works on current
processors. On the older stuff, there were two multiplier values.
AMD has multiple P-states for their processors, but I don't know
if Intel does now as well or not. I cannot find any details on
the Intel site, for how many P-states they use. But that might be
why the multipler value is changing. Disabling EIST in the BIOS
might be the first step I'd try, to make it stop.

Paul

>
> <tom RemoveThis @japtuning.ch> wrote in message
> news:1172776523.511628.272630@t69g2000cwt.googlegroups.com...
> | And does somebody nows something about this one?
> |
> | The other thing is, when I look with CPU-Z is my multiplier switching
> | himself from 9x to 6x and back...
> |
>
>
 >> Stay informed about: E6600 and D975XBX2 

'Paul' wrote, in part:
| It could be EIST.
|
|
http://www.intel.com/cd/channel/reseller/asmo-na/eng/products/box_proc...ors/des
|
| I haven't been able to find any info on how EIST works on current
| processors.
_____

The behavior seen by the original poster IS Enhanced Intel Speedstep
Technology at work.

Additional information is at
http://www.intel.com/support/processors/core2duo/
and
http://www.intel.com/design/processor/datashts/313278.htm
(particularly 6.3)
and
http://www.intel.com/design/processor/specupdt/313279.htm
and
http://www.intel.com/cd/channel/reseller/asmo-na/eng/203838.htm .

Phil Weldon





"Paul" <nospam.RemoveThis@needed.com> wrote in message news:es8c5d$fl3$1@aioe.org...
| Phil Weldon wrote:
| > 'Tom' wrote:
| > | And does somebody nows something about this one?
| > |
| > | The other thing is, when I look with CPU-Z is my multiplier switching
| > | himself from 9x to 6x and back...
| > _____
| >
| > My answer applies to the Intel E6600. What you see with CPU-Z is
'thermal
| > throttling'. The CPU, in a completely on-CPU-chip action that REDUCES
the
| > CPU multiplier to lower heat production if the CPU temperature becomes
too
| > high. You can disable 'thermal throttle' (automatic CPU multiplier
| > reduction [based on CPU temperature], but you can NOT directly change
the
| > CPU multiplier. You can NOT INCREASE the CPU multiplier.
| >
| > Phil Weldon
|
| It could be EIST.
|
|
http://www.intel.com/cd/channel/reseller/asmo-na/eng/products/box_proc...ors/des
|
| I haven't been able to find any info on how EIST works on current
| processors. On the older stuff, there were two multiplier values.
| AMD has multiple P-states for their processors, but I don't know
| if Intel does now as well or not. I cannot find any details on
| the Intel site, for how many P-states they use. But that might be
| why the multipler value is changing. Disabling EIST in the BIOS
| might be the first step I'd try, to make it stop.
|
| Paul
|
| >
| > <tom.RemoveThis@japtuning.ch> wrote in message
| > news:1172776523.511628.272630@t69g2000cwt.googlegroups.com...
| > | And does somebody nows something about this one?
| > |
| > | The other thing is, when I look with CPU-Z is my multiplier switching
| > | himself from 9x to 6x and back...
| > |
| >
| >
 >> Stay informed about: E6600 and D975XBX2 

Phil Weldon wrote:
> 'Paul' wrote, in part:
> | It could be EIST.
> |
> |
> http://www.intel.com/cd/channel/reseller/asmo-na/eng/products/box_proc...ors/des
> |
> | I haven't been able to find any info on how EIST works on current
> | processors.
> _____
>
> The behavior seen by the original poster IS Enhanced Intel Speedstep
> Technology at work.
>
> Additional information is at
> http://www.intel.com/support/processors/core2duo/
> and
> http://www.intel.com/design/processor/datashts/313278.htm
> (particularly 6.3)
> and
> http://www.intel.com/design/processor/specupdt/313279.htm
> and
> http://www.intel.com/cd/channel/reseller/asmo-na/eng/203838.htm .
>
> Phil Weldon
>

What I'm referring to, is when they say:

"Multiple voltage/frequency operating points"

they don't say how many points. EIST on older processors was
noteworthy, because it used precisely two points. I'm curious
if Intel has invested all the extra effort in VRD11 and
D-VID, only to still have the same two operating states.
AMD was different, because they test their processors at
more than two P-states, but Intel provides no such
explicit information about their products. Makes me
wonder how the BIOS knows precisely what ACPI objects
to pass to the OS (i.e. where is the knowledge about
the number of FID/VID combinations stored). At least
with AMD, you could look up what states were supported.

Paul
 >> Stay informed about: E6600 and D975XBX2 

I did disable EIST, but it is still switching between 6 and 9, any
other idees?
 >> Stay informed about: E6600 and D975XBX2 

Phil Weldon wrote:
> 'Paul' wrote, in part:
> | they don't say how many points. EIST on older processors was
> | noteworthy, because it used precisely two points. I'm curious
> | if Intel has invested all the extra effort in VRD11 and
> | D-VID, only to still have the same two operating states.
> | AMD was different, because they test their processors at
> | more than two P-states,
> _____
>
> Intel historically has produced more robust CPUs than AMD. Intel EIST is
> implemented on-CPU-die for thermal control, one step-down seems sufficient,
> and clearly is the explanation for the original poster. What purpose would
> additional CPU multiplier and CPU core voltage setbacks serve? Other than
> to offer operation using variously inadequate thermal solutions?
>
> Phil Weldon
>

EIST is for power savings, not for thermal reasons. If the processor
is not being used, it drops to a low power state. And has the ability
to return to the higher power state, in very little time.

Thermal Monitor is for thermal control, and involves clock throttling.

"When the Thermal Monitor feature is enabled and a high temperature
situation exists (i.e. TCC is active), the clocks are modulated by
alternately turning the clocks off and on at a duty cycle specific
to the processor (typically 30-50%)."

With regard to EIST, AMD does the same thing, only AMD is load sensitive
with more power states. I'm not saying that is particularly beneficial, but
it does allow AMD to run the CPU closer to 100% relative load, by choosing
one of the intermediate P-states. I guess I've seen very little investigation
into whether those features actually work or not, and Intel not giving
details just makes matters worse.

Both companies have their faults when it comes to documentation. I've
noticed, since AM2 came along, that AMD has become very lazy, when it
comes to documenting their products. The kind of info they offered for
S754 and S939, is missing for AM2. In fact, I don't think I can find
P-state info for AM2, just for earlier processors. There isn't even a
generic AM2 socket spec. It is almost like they fired the tech
documentation department. The only doc I did find that was useful,
was the BIOS writers guide for AM2 (32559.pdf).

Paul
 >> Stay informed about: E6600 and D975XBX2 

WHERE did you disable EIST? There is a BIOS setting and a Windows XP
setting. See the reference from my previous post
http://www.intel.com/cd/channel/reseller/asmo-na/eng/203838.htm ,

under 'Integration'
for the BIOS:
1.. Ensure Enhanced Intel SpeedStep technology is enabled in your BIOS.
For Intel desktop boards the Intel SpeedStep technology option is under the,
"Power," tab and labeled, "EIST." Ensure it is set to, "Enabled."
for Windows XP:

Under Control Panel - open Power Options
a.. Under the Power Schemes pull down menu
a.. To turn Enhanced Intel SpeedStep technology ON select, "Minimal
Power Management," power scheme.
b.. To turn Enhanced Intel SpeedStep technology OFF select, "Always On,"
power scheme.
More than that I cannot say, but only recommend you contact Intel customer
support.

Phil Weldon

<tom.RemoveThis@japtuning.ch> wrote in message
news:1172868101.337945.162660@t69g2000cwt.googlegroups.com...
|I did disable EIST, but it is still switching between 6 and 9, any
| other idees?
|
 >> Stay informed about: E6600 and D975XBX2 

'Paul' wrote, in part:
| they don't say how many points. EIST on older processors was
| noteworthy, because it used precisely two points. I'm curious
| if Intel has invested all the extra effort in VRD11 and
| D-VID, only to still have the same two operating states.
| AMD was different, because they test their processors at
| more than two P-states,
_____

Intel historically has produced more robust CPUs than AMD. Intel EIST is
implemented on-CPU-die for thermal control, one step-down seems sufficient,
and clearly is the explanation for the original poster. What purpose would
additional CPU multiplier and CPU core voltage setbacks serve? Other than
to offer operation using variously inadequate thermal solutions?

Phil Weldon

"Paul" <nospam.TakeThisOut@needed.com> wrote in message news:es8jd5$u7c$1@aioe.org...
| Phil Weldon wrote:
| > 'Paul' wrote, in part:
| > | It could be EIST.
| > |
| > |
| >
http://www.intel.com/cd/channel/reseller/asmo-na/eng/products/box_proc...ors/des
| > |
| > | I haven't been able to find any info on how EIST works on current
| > | processors.
| > _____
| >
| > The behavior seen by the original poster IS Enhanced Intel Speedstep
| > Technology at work.
| >
| > Additional information is at
| > http://www.intel.com/support/processors/core2duo/
| > and
| > http://www.intel.com/design/processor/datashts/313278.htm
| > (particularly 6.3)
| > and
| > http://www.intel.com/design/processor/specupdt/313279.htm
| > and
| > http://www.intel.com/cd/channel/reseller/asmo-na/eng/203838.htm .
| >
| > Phil Weldon
| >
|
| What I'm referring to, is when they say:
|
| "Multiple voltage/frequency operating points"
|
| they don't say how many points. EIST on older processors was
| noteworthy, because it used precisely two points. I'm curious
| if Intel has invested all the extra effort in VRD11 and
| D-VID, only to still have the same two operating states.
| AMD was different, because they test their processors at
| more than two P-states, but Intel provides no such
| explicit information about their products. Makes me
| wonder how the BIOS knows precisely what ACPI objects
| to pass to the OS (i.e. where is the knowledge about
| the number of FID/VID combinations stored). At least
| with AMD, you could look up what states were supported.
|
| Paul
 >> Stay informed about: E6600 and D975XBX2