Another speedstep problem...but worse!?

Hey guys, now i know you have all heard about people having this problem, having the cpu speed capped off due to speedstep, but after hours of googling, no one else seems to have my problem.

I'm currently using a Dell Inspiron 9200, with a pentium M @ 2.10ghz. Just recently i noticed that it was only running at 598mhz. I learned about speedstep, and tried a couple things. Most people report that when disabling speedstep in bios, frequency is locked at half of potential. I know my math sucks, but when i disabled speedstep, it was still locked in at 598mhz, which im sure is not half of 2.10ghz.

I also tried changing power settings/profiles, using speedswitchxp, and trying other different combos. Then I ran across flipfire's guide to "undervolting." Installed HWmonitor, Orthos, and RMclock. Now, i haven't changed any settings on RMclock yet, because I did the Orthos temp check first.

With speedstep enabled and with power profile set at always on (and speedswitchxp uninstalled), I tried the heat check for 10mins. Core clock and throttle stay at 598mhz even with OS load and CPU load maxing out. Temp after 10 mins was recorded at 37c/97F. And once again, speedstep IS enabled.

So what to do? If I have missed anything, please let me know. As a gamer, running on 600mhz makes me sad, knowing it has a potential of 2.1ghz. Thanks in advance!

 

Did you do this w/speedstep enabled ?

You can also try to fix a particular multiplier with CrystalCPUID.

 

First of all, how are you checking the system so that your sure it's 598mhz?
Speedstep technology is a "performance - on - demand" technology, so in other words you will need to run programs that would require 100% processing power for it to ramp up to full speed. If all your doing is using utilities to monitor your processor speeds while the cpu is pretty much idle, you'll never see it run at maximum.

Use stress tester like Prime95 / Orthos or benchmark software (like superPI). To be honest, you won't want it to run full speed all the time, the amount of power require and heat it would create would be undesirable.

As far as running games, I'm pretty sure you're using all 2.1Ghz or else your games would run very sluggishly.

 

My first post mentions that I did use Orthos. I had it running for 10 minutes straight, as suggested by flipfire's guide. And yes, my games do run sluggishly--I have to run all settings at its lowest, and even then I still lag.

I'm going to try it with speedswitchxp and see what happens..

 

If that doesn't work^, do a reg scan with CCleaner, to get rid of any keys left by SSXP.

And also run the RMClock_WipeOut.reg, and try the guide again. And also try disabling all Advanced CPU options in RMClock.

 

Thanks to dex and andy posting replies.

Andy, I tried cleaning the reg, ran wipeout.reg again, and it's still stuck at 598mhz during Orthos test. I also tried disabling advanced CPU options.

I just want to point out that I haven't tried changing advanced CPU options, via step 3 and onward on the guide, because I assume that I should get normal results first. Or should I try the rest of the guide anyways?

Thanks again guys.

 

Yeah, try out the rest of the guide. UVing ain't bad.

Also try out CrystalCPUID, and see if you can fix a particular multiplier.

 

Well, even though the initial stress test failed again, changing the rest of the settings on RMclock worked! Thanks a bunch for your help, I can play in peace now!

 

Which settings did you change?

 

Sorry mnyugen, misread some of your original post. I'm glad the rest of the guys sort of solved your issue

 

I know this thread is kind of old, but I ran into the exact same issue. My 2ghz T7250 is currently locked to multiplier 6 no matter what and is running at 1,2ghz. I ran the rmclock wipeout.reg once and tried cpugenie as well as neither, but it stays at 1,2. I once spotted it going to multiplier 8 but couldn't get that again.
And yes, performance is affected a lot. SuperPI is over 40 secs now. What rmclock settings did you change ?

Only thing I can achieve with the advanced settings is that it will throttle down to superLFM, but at 100% load it just goes to normal 6x. EIST setting is greyed out (and checked) in rmclock.

 

This is starting to get on my nerves. No solution that I've found has worked. The CPU is locked at 6x multiplier no matter what, rmclock cannot change anything above that, but voltage changes at multiplier 6x works.
Crystalcpuid even showed that my cpu would clock down to like 250mhz when idle which I don't really believe.

 

Have you checked your temps?

 

I think you lowered the voltages of the multipliers other than your max multiplier's voltage. If I'm right, then maybe you forgot to re-tick the multiplier boxes in the "performance on demand" sub-profile.

A quote from flipfire's guide:

" Note: After you're done, make sure you re-tick the multiplier boxes in the "Main Profile" page then "Performance on Demand" sub-profile page, If you don't your CPU will be majorly handicapped."

 

Sorry, but its not that simple Im not using rmclock for the 1st time and I know what does what. Temperatures are and have been way below anything harmful.

The problem persists through 2 formats (vista and wisdows 7) with NOTHING installed. Right after install I started cpuid and orthos and tested superPI, its still downclocking on AC. Even tried reflashing bios.
Even tried recalibrating battery, though not sure what it's supposed to do. Im running out of ideas here.

On AC: The CPU idles at 8x multiplier and 1,6ghz and 0,9V and on full load 6x multiplier and 1,0V.

On Battery: Everything is normal. Multiplier shoots to 10x the moment I unplug AC.

 

Just to make sure you know, if Speedstep is disabled, the CPU will be locked at it's lowest multiplier. Of course if Speedstep is enabled, then it isn't what's causing your problem.

 

Freakin sweet! I had thought about the AC not supplying enough power, but didn't know what to do about it.
My suspicions just got worse when I noticed that above 58% max processor usage the FSB dropped by 10! I though that the machine isn't getting enough power as the same happened when I plugged the battery in (some power to charging).

Well, I unplugged everything, uninstalled the wifi card and the other memory and booted on battery. You can't imagine how relieved I was when plugging in the power on full load this time the multiplier stayed on full

Now, as I put everything back as they were, its on full performance again on AC. Phew. I don't know what happened, but something must have reset about the AC supply power, because its working now and everything is exactly the same as when it wasn't working. So far so good..

edit: I take that back.. it reverted to 6x once again on AC

 

Sigh, its back on 6x no matter what I unplug or disable.
I remember noticing earlier that on AC power one of the usb ports leaks power, could this be connected ?

Im not even sure if this happened after or before I set my fsb to 210 with setfsb. I noticed it after that, anyway, wondered why wouldn't my multiplier lock to 10x with rmclock.
And it hadn't happened before with setfsb, but I haven't really even played with setfsb.

 

This is getting stranger by the minute. I noticed that the voltage and multiplier my CPU downclocks to is the Thermal monitor 2 target (which is basically when a cpu reaches 100C or 125C and it downclocks to those numbers, which hasn't happened 100% sure).

I didn't even think I could disable thermal monitor 1 and 2. I just unticked those 2 from rmclock and BAM!, cpu shoots up to 10x on AC. Just praying it actually stays there this time.
Oh, and it didn't disable temperature monitoring.

edit: To be precise, disabling only (Automatic thermal protection) Thermal monitor 2 worked. 1 Is still active and you bet im not gonna let my CPU cook because i've disabled the TM2. Im sure it was just a bug somewhere, maybe even on hardware level.

 

Okay, had to disable TM1, too. I hate to say it, but having to disable all of CPU's automatic thermal protection to get it working full speed is not far from suspicious. But I can't figure out why would the TM1 and TM2 be working like this as every single temperature monitor on the machine says everything is OK.