i7 Slower after overclocking with turbo boost

Hey everyone, i have been trying to overclock my m11xr2 and i am pulling my hair out over this problem. When TB is disabled raising the fsb in the bios raises my cpu score in the bios. However if i raise the fps with turbo boost enabled the performance is the same/slightly worse than at the stock fsb. Here are my 3dmark06 cpu scores so far:

OC Disabled TB Disabled 1333
OC 160mhz TB Disabled ~1600 (I can run it again for exact score if needed)
OC Disabled TB Enabled 1784
OC 166mhz TB Enabled 1759
OC 146mhz TB Enabled 1728

Specs:
m11x r2
i7 Stepping 5 Rev K0
4gb ddr3
Seagate momentus XT 500gb ssd/hdd hybrid
GT 335m 543/1303/905 258.96
Intel HD driver 8.15.10.2189
Clean installation of windows 7 Ultimate x64

If anyone has any suggestions that would be great. I am trying to break 8k and based on other peoples R2s i should be able to get 2k or near 2k cpu score with that OC.

Thanks in advance.

 

A small update. I did a full 3dmark06 run at 160mhz and the cpu score was even lower(maybe from the heat of the graphics test? :/ ).
I installed the intel 8.15.10.2202 drivers.
There are no processes using up a bunch of cpu power and i turn off any chat programs ect. before benching. Also no antivirus or similar software is installed or running.

Here is a pic of the last run including a cpuz window ect.
http://img521.imageshack.us/img521/1636/m11bench.jpg


btw is there an edit button somewhere so i don't have to keep posting replies?

thanks

edit: found it =P I have no idea how i missed it before.


edit2: I brought the computer outside were it is ~13c keeping it plugged in and got:
3607
3631
1771

The system was cool enough i don't think it ever went past the first fan speed so that rules out heat.

 

IS it affecting any SW you are running (Besides benchmarking SW)

IF not just forget about it and enjoy the lappy.....People get too hung up in "benchmark numbers"....Real world is what counts.

 

Everything is running ok but if there are performance improvements to be had i will go for them. Allot of the enjoyment i get out of computers is benching and finding ways to make them go faster. But this is really bugging me since at my current settings i should be reaching 8k, even some with the i5 version are getting better cpu performance than i am =/

 

Have any of you guys disabled CORE parking in Windows 7? This link shows the simple way to disable core parking and may give you a more "snappy" processor response.

Windows 7 & Core Parking .... a *better* way to Turn It OFF ....

Another thread to explain...
http://ultimatecomputers.net/forum/viewtopic.php?f=7&t=3644


Best Wishes, StevenX

 

Learn something new every day. Thanks, StevenX!

 

what about 166 with TB disabled?

 

Adding Value.. it's what i do... Thanks Slickie!

 

For anyone who tries disabling core parking, please let us know if it interferes with Turbo Boost, i.e., if TB happens less often or not at all or not to the highest multiplier.

My understanding of Turbo Boost is that Windows schedules most of the threads to the first core in order for it to activate Turbo Boost to a greater degree, even though scheduling more to one core than the other seems to contradict the whole idea of having more than one core. This is because TB won't happen unless the CPU is at 100% usage, so in order to be effective, it must keep the core working and maxed out. Also, TB kicks in higher when additional cores are disabled. Now, weigh both of those facts against the purpose of Core Parking, which is to disable the second core so that the first processor is used at a higher % (e.g., instead of 10% usage on both cores, the second is disabled so that the first is running at 20%--which saves power and activates TB of the first if the combined % is >=100%), which is exactly what needs to happen for Turbo Boost to work right. If the additional core isn't disabled, then it would seem TB on the first core won't up the multiplier as high on the first at the times when there is a full load at the base frequency on one core.

Perhaps Core Parking isn't the only feature affecting the second core so that disabling it doesn't affect TB at all; I really don't know, which is why I asked

edit: found this thread on some people having the same discussion Win 7's Ideal Core/Core Parking - Beyond3D Forum though I don't really see a consensus.

 

@Corwin - good points. I have had this disabled on mine for months but I wanted to see how the I series will respond to it... Again if you dont like it, just reverse the entry in the registry. No lasting harm done!

You can always run some before and after game benchmarks to see how impacts you.

BW, StevenX

 

I'll try this tomo, sounds quite interesting.

Thanks to original poster.

 

Turbo Boost is only activated if the chip thinks its temperature is low enough to do so, so I suspect that raising the FSB is making it run hotter, so it doesn't boost (or does so less often) so overall it isn't as quick. Perhaps try it with some additional cooling for the laptop?

 

Thanks for all the replies everyone. I don't think it is heat since i tried benching outside in 13c temperatures. The laptop won't post at 166FSB and is semi stable at 158fsb. I have backed it off to 150fsb and am getting the same performance.

I will play around with that core parking thing tomorrow and update you guys on the effects. Any specific things you want a before or after of before i make the change?

 

I set it to disable that feature on AC power and i can confirm it worked using performance monitor. I have not noticed much of a difference so far. I benched 3dmark06 again and got 3626 3637 1718 so no notable difference there. I will try some games tomorrow and see if i notice anything.

Since it doesn't seem to hurt performance it is probably worth disabling for AC power but pointless to disable it on battery.

 

It would appear disabling core parking is mostly if only beneficial when gaming.

I've just noticed a rather large increase in FPS when playing COD4. Peaks of 80 - 90. I always keep an eye on Fraps and have never noticed numbers in that range prior to editing the registry.

The multiplier doesn't peak as much from looking at CPUZ.

 

You know what I just noticed... Your i7 is stepping 5, revision K0. My i7 is stepping 2, revision C2. Very interesting...

 

K0 should be more overClarkable Mine's a C2 unfortuneately

 

If disabling Core Parking affects Turbo Boost, it would do it when there is enough of a load to "boost" one core and not the other. When benchmarking or anything that's maxing out the cpu, TB is going to boost both to the max and Core Parking if on won't disable one core, and it obviously won't disable one when off, so it shouldn't be any different. In other words, it should be day to day usage--light to moderate usage--that changes, if anything.

 

my K0 is a very poor overclocker. It won't even post at 166 and needs to be near 150mhz to prevent a BSOD when turbo is high for a long time.

 

Exactly the same with my K0 too.

 

Rev. C2 overclocked @ 158MHz......no problem at all

 

Does anyone think I would have any luck getting dell to do a board replacement? I know there is nothing technically malfunctioning but my I spent money on an I7 that in many cases performs the same or less than an I5.

Then again I'm not sure i want to gamble getting back a unit with hinge issues ect. Though my keyboard does touch the screen when it's closed and my keyboard leds may be dimmer than they should(not sure how to test that).

edit: I don't really understand how the number/letter system in cpu stepping works. Are the C2's a newer/higher stepping than K0?