speedstep does not work after upgrade to XP on a R30

[doctor_faustus]

Dear thinkheads,

Under Win2000 I had the problem that speedstep was doing exactly the opposite of what was expected: Fast at the low-fequ setting and vice versa. I could live with that.

Under XP the only way to influence the processor speed is to disable speedstep in the BIOS entirely: then the CPU runs at low frequ. All other modes in the BIOS (Automatic and always full power) cause the CPU run at high frequency all the time. I have tried all power schemes of IBM's batterymaximizer utility and also other speedstep control programs like the freeware speedswitchxp1.4. I was not able to influence the speedstep mode in a single case independent whether running on battery or AC. I have measured the actual processor frequency with two independent tools.

Is anyone using a R-class thinkpad under XP with actually working speedstep or does anyone have an idea what might be wrong?

[Riddil]

XP has it's own built-in speed-step. In 2K you needed a utility to get it to work right. I don't know, but I suppose it's possible that if you have the old 2K speedstep still running that it might interfere with the XP built-in speedstep.

Try uninstalling it, and then go to the ibm driver download and install the latest power management driver. I don't use the battery maximizer for much than monitoring battery charge, but when I use "Portable/Laptop" power scheme from the Windows "power" control panel it works fine for me. It idles at the low power state when CPU usage is low, and then jumps to max when CPU usage goes high.

[doctor_faustus]

Thanks for your note! I tried to locate any remnants of the original Intel utility without success so far. I think I read somewhere that XP is quite efficient in killing it.

I have since observed the behavior your are describing: If the CPU idles the clockrate goes down. However, that's of little use for me because if the CPU idles its power consumption is down anyway. If I need to save significant amounts of energy (prevent overheating in my case) then I would rather use a low clock rate even at full CPU load. Thats how it worked under win2000 when, e.g., on battery.

More precisely asked: Is anyone able to reduce the clock rate deliberately (without changing the BIOS setting), if only to prevent the (noisy) fan from turning on?