Core Parking!

[Virgorising]

The way I do it is in the registry. Open regedit, navigate to HKEY_Local_Machine > System > ControlSet001 > Control > Power. In the right hand window there is a key CoreParkingDisabled. Default is 0 (off), change it to 1 (on). Repeat for ControlSet002 and CurrentControlSet. Exit regedit and reboot.

You mean you do this to disable parking, yes? Well I just used the most benign app and it worked, but thanks for this. Good to know ALL THE ROADS TO ROME!!!

 

[krose]

Yes, to disable parking. I prefer to do things myself rather than rely on third party software of perhaps questionable integrity. You can easily write two reg files, one to turn parking off and another to turn it on. Open notepad. Copy and save as coreparkingoff.reg

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\ControlSet001\Control\Power]
"CoreParkingDisabled"=dword:00000001

[HKEY_LOCAL_MACHINE\SYSTEM\ControlSet002\Control\Power]
"CoreParkingDisabled"=dword:00000001

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Power]
"CoreParkingDisabled"=dword:00000001

Change the dword value 1 to 0 and save as coreparkingon.reg. Just double click a file to change the setting, then reboot.

 

[Virgorising]

Yes, to disable parking. I prefer to do things myself rather than rely on third party software of perhaps questionable integrity. You can easily write two reg files, one to turn parking off and another to turn it on. Open notepad. Copy and save as coreparkingoff.reg

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\ControlSet001\Control\Power]
"CoreParkingDisabled"=dword:00000001

[HKEY_LOCAL_MACHINE\SYSTEM\ControlSet002\Control\Power]
"CoreParkingDisabled"=dword:00000001

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Power]
"CoreParkingDisabled"=dword:00000001

Change the dword value 1 to 0 and save as coreparkingon.reg. Just double click a file to change the setting, then reboot.

I respect yr meticulousness and standards.

Now, just curious, What OS are you running, and, what improvements do you find having disabled core parking?

 

[Virgorising]

http://support.microsoft.com/kb/2646060/EN-US

Just found the above. It appears it does not apply to Intel chips, yes?

Boy, this is complex stuff, the core parking, the debates, the CPU-specificity....

 

[ShintaiDK]

http://support.microsoft.com/kb/2646060/EN-US

Just found the above. It appears it does not apply to Intel chips, yes?

Boy, this is complex stuff, the core parking, the debates, the CPU-specificity....

Be careful, it was pulled once and made into a request hotfix rather than patch due to BSOD.

 

[Virgorising]

Be careful, it was pulled once and made into a request hotfix rather than patch due to BSOD.

Thank you, yes! I am being careful verging on paranoid!!!! Which Is why I am posting.

Plus, what I came upon seems only to address AMD chips.

I turned parking on again. During the interval when it was turned off, my take was, the system was not as "happy" as it is with it turned on. I have no official, benchmark level data, cept the subjective test doing video encoding and my overall sense......but this system just seems faster and more stable with the parking turned on.

Do I understand this?

NO.:| But, as always, I am TRYING TO.

 

[Phynaz]

So I just did some playing around with this.

At idle it makes no difference.

At somewhat heavy load load it increased performance by a few percent with a few percent increase in power consumption. I'll post more with full load numbers.

 

[Virgorising]

So I just did some playing around with this.

At idle it makes no difference.

At somewhat heavy load load it increased performance by a few percent with a few percent increase in power consumption. I'll post more with full load numbers.

Wow! Isn't THIS interesting!?

May I know what OS you are running and if the chip in question is Intel?

 

[Phynaz]

Windows 7 64 on an Intel i7-870.

Full load shows no difference with core parking turned off.

I might as well leave it off.

 

[Virgorising]

Windows 7 64 on an Intel i7-870.

Full load shows no difference with core parking turned off.

I might as well leave it off.

Interesting! I HAVE THE SAME CHIP! Lynnfield! i7 870 2.93....and also running same OS.

As of today, I think I'll leave mine on. But I am going to remove it again to see if my impression of the system with it off changes.

Tell me, what did you do to test things? My only CPU-intensive activity is video encoding once in a while, so, when I tried to test, I din know how to bring the chip up to full load. Never saw this chip at full load.

 

[Phynaz]

I tested with Winrar, 7Zip, Cinebench and TMPGEnc Video Mastering Works.

 

[Virgorising]

I tested with Winrar, 7Zip, Cinebench and TMPGEnc Video Mastering Works.

I am impressed. I never even heard of the last three!!!!!

 

[Phynaz]

I spend more time in TVMW than any other apps. I've done 40 hour batch encodes

 

[Phynaz]

Some more info from the Anandtech article:

http://www.anandtech.com/show/7189/choosing-a-gaming-cpu-september-2013/2

Wow, I completely missed that article. Thanks for the link.

 

[atm743]

http://support.microsoft.com/kb/2646060/EN-US

Just found the above. It appears it does not apply to Intel chips, yes?

Boy, this is complex stuff, the core parking, the debates, the CPU-specificity....

Currently, the CPU Power Policies that are used by Windows 7 and Windows Server 2008 R2 are not optimized for the dual core AMD Bulldozer module architecture. This architecture is found on AMD FX series, AMD Opteron 4200 Series, and AMD Opteron 6200 Series processors. This can result in decreased system performance with multithreaded workloads in lightly-threaded environments.

the update is for these cpu's

download http://bitsum.com/about_cpu_core_parking.php works great.

 

[Tsavo]

The last time I parked my cores, someone stole my hubcaps.

No more core parking for me.

 

[Phynaz]

Okay more tests...

Using MultiPar I took two 700MB files and created a 860MB power of two repair set. Then I deleted one the of the files and ran a repair. Files were cached in memory for read, writes were to an SSD. I did each test twice. Times are min:sec Note Multipar will use a max of four logical cores.


Thread parking on, create
4:24
4:27

Thread Parking off, create
4:25
4:25

Thread Parking on, repair
3:42
3:39

Thread Parking off, repair
3:38
3:41

Outcome - No difference in this app.

 

[Virgorising]

The last time I parked my cores, someone stole my hubcaps.

No more core parking for me.

:biggrin::wub:

 

[Rubycon]

The last time I parked my cores, someone stole my hubcaps.

No more core parking for me.

Try docking cores, it's much more resistant to strong blows.

 

[Virgorising]

Try docking cores, it's much more resistant to strong blows.

Forgive my not knowing, but I jus don't, is this a cute joke or a real phenomenon? The docking, I mean.

I am even more overwhelmed/confused by core parking everything now than when I first learned it existed.:\

Would be amazing if someone who works at Intel could post here!!!

 

[Rubycon]

Forgive my not knowing, but I jus don't, is this a cute joke or a real phenomenon? The docking, I mean.

I am even more overwhelmed/confused by core parking everything now than when I first learned it existed.:\

Would be amazing if someone who works at Intel could post here!!!

Just joking for sure.

There are a few Intel employees that are posters here, perhaps they may chime in.

 

[Virgorising]

Just joking for sure.

There are a few Intel employees that are posters here, perhaps they may chime in.

Sigh. I am so overwhelmed with confusion re core parking, I actually did not get it.

I am stunned and heartened there are actual, official Intel pundits here! Where are they on this issue and how can we give them a shout out?
___________________________________
Edit: I feel near masochistic that I keep doing searches on this thing.....I just now found this, whose content is way above my pay grade.

http://forum.cakewalk.com/Core-Parking-m2628147-p3.aspx

 

[Phynaz]

Nobody from Intel is here in any official capacity. They are just regular forum members who happen to be employed by Intel.

If they have something to say about this personally they will post. They will never post in any official capacity or speak as representatives of Intel.

 

[tweakboy]

Just be careful you dont get a ticket.

 

[Virgorising]

Nobody from Intel is here in any official capacity. They are just regular forum members who happen to be employed by Intel.

If they have something to say about this personally they will post. They will never post in any official capacity or speak as representatives of Intel.

I can understand that from a corporate protocol point of view. I could not care less if they keep their "INTEL INSIDE" TATS covered up:sneaky:.....cause they must all have that ink, right?() I just want them to show up.

AMD pundits also welcome, for sure, but they gotta get in line.

I would add MS pundits, but I am not a cretin.:sneaky:

My take now, is we need something official in this convoluted journey. An actual trial with expert witnesses offering hard data, being cross examined, etc. and we all stream it. But, to pas voir dire, no jury member need be MS certified. In fact, that would actually be a deal breaker.:whiste: