G50 FSB/multiplier trottling

Hi,

I know the multiplier of my P8600 switches from 9 to 6 on battery mode, and I remember seeing it at 10 (confirmed by someone here too) in certain cases.

Now, I see it trottling from 9 to 9.5 very often (5 times / minute), and in the same time, the FSB often trottles down (from 266 to 252 MHz).
Sometimes, the FSB doesn't trottle down, resulting in higher CPU freq : 2400 to 2527 in normal mode.

I've been confirmed on the french board that it happens in XP and Vista, and other users see it too.
I see it using CPUz and Everest.

Could someone here please explain this ?

 

Its the new Intel chip set managing your fsb and its actually great. When your computer doesn't need the power it drops the multiplier so it uses less power and the split second you need the processing power it bumps it back up again. Even if you don't care about battery life it keeps your laptop cooler and thus will last longer. Don't worry as you have full power when you need so just go about your daily business and forget about it!

 

c-states. if im reading it correctly, then see if it still happens if you disable C1 in RMclock.

 

Where can I find this c-state setting ? I can find P-state, but no 'c'...
I tried to change some settings here and there in RMclock, but nothing seems to work.
Maybe there's a conflict between NHC and RMclock ?

I don't complain about the FSB trottling, I'm wondering about the CPU mult. going to x9.5, even x10, when it's designed for x9 max...
I see my CPU jumping from 2.4 to 2.7 without overclock and without high usage, so I'm surprised...

 

Advanced CPU settings ->, theres Intel ... low power states. try it after having all unchecked.

 

Thanks Thalanix.

You can see my x9.5 multiplier, resulting in my CPU going from 2400 to 2526, without doing anything (computer is idle during the screenshot)

As you can see, it's still happening, even with lowpower states disabled...
How can I overclock my CPU right when I can't even fix a multiplier ?

 

Just a shot in the dark here but could having the DFFS (Dynamic FSB Frequency Switching) enabled be causing this? I see this on the left window of your screen shot near the bottom. I'm not sure what it is hence the "shot in the dark" but it sounds like it might effect things this way.

 

i don't think it will change much, but see if it reads the registers differently if you switch the type to Mobile (in the screenshot).

 

Disabling DFFS or switching to mobile mode don't do anything
I also tried to uncheck everything in low power states settings in mobile mode, but I see no difference.

Disabling DFFS seems to actually stop FSB trottling, but that's all.

On an other side, adjusting vCore or other settings don't seem to do anything, so maybe the chipset is not supported by RMclock ?

 

are you sure cpu-z is reporting it correctly?
it sounds similar to the two screenshots ive seen of the t9400 going to a 10x multi, which if it did, the math didn't make sense (and both cpu-z and setfsb reported a different fsb speed). but that was after overclocking with directconsole.

also, see what happens if you disable Thermal Monitor 2 and Extended Throttling.

 

I think u're right as RMClock and Cpu-z are not reporting the same informations (getting a shot of this is quite tricky ^^)

So I can assume it was a reading error, and cpu never got to x9.5 ?

 

rmclock doesnt support half-multipliers. would say core 2527/throttle 2398 if it was at 9.5. but yea, it looks as if its being read wrong.

 

Ok, so thanks

Any news about fan speed bios mod ? I just have read your last post one more time, but I couldn't help you with that, I only understand 1/2 words

 

i put it on hold until after exams (last on on the 27th). basically, its just about finding when the fan speeds up, then finding those temps (albeit in hex) in the dsdt file i posted.