CPU Undervolting problem

I have a ThinkPad T61p with a T7700 CPU. I've followed the undervolting thread and the PHC articles in the wiki, but when I execute "linux-phc-optimize", the script lowers the VID all the way down to 0, but the system never crashes. Someone suggested that this could actually be a good thing, that my CPU can work with only the base voltage (~700 mV) at the highest frequency. That sounded too good to be true.
So I've tested it with echoing half the default values to /proc/sys/something/.../cpu0/... and the system completely froze. The picture remained on the screen, but I've lost all control. It needed a hard reboot and some fsck's for fs corruption.
So, what do you think is wrong? Any suggestions?

Dont use linux-phc script. I think it uses only one core, so it is not good for C2D. Rather lower freq step by step. And use your lap, as you do everyday. If you notice lags/slowdowns/fs errors you know you find your proc maximum low.

Similar Messages

  • My solution (or so I thought) to Power Manager/CPU throttling problems!!!

    I have found a solution to all my Power Manager/CPU throttling problems!!! Though there is some good and bad news.
    {EDIT: The problem has NOT been fixed, even after latest PowerManager (3.62) and BIOS (1.30) versions. Pretty much ignore anything I say below as the problem is still occurring. You can see my full post here: http://forum.lenovo.com/t5/W-Series-ThinkPad-Laptops/W520-Speedstep-not-working-properly-on-battery-...}
    Good news:  I have NONE of the throttling issues or inconsistent CPU frequency problems I was having before on AC or battery power.  Everything, including TurboBoost on battery works! It is completely fixed! (I have no idea how this factors into Lenovo’s statements that TurboBoost is disabled on battery “by design”. There is at least one other post from someone else that also reported TurboBoost was working for them on battery)
    Bad news: I don’t really know which one of the many things I tried actually worked. I am sorry I wasn’t more methodical about recording what I did and checking results, but this was my last ditch effort to get this fixed on my own without sending the system in for repair and frankly, I didn’t think it would work. Now that it has worked, I’m hoping my steps can help others.
    For anyone interested, here’s what I did… and before anyone says something like “That has nothing to do with managing power/cpu, why would that help?!… etc., please keep in mind I’m just stating exactly what I did. I am aware some of the steps may not be relevant, but who knows… We all know how weird PC’s are sometimes, even the smallest, oddest thing may resolve a problem.So anyway, here goes. 
    **IMPORTANT** Not sure how many noticed, but there was a new version of Power Manager released a few weeks ago, 3.62. The PM driver seems to have stayed the same. That alone could very well be the sole fix, I’m not sure. You may just want to completely remove PM and PM driver and install the latest version before trying any of the steps below.
    1)Made a complete system image via Windows built-in backup feature
    2)Disable any 3rd party fan/CPU control utilities (Throttlestop, etc). Make sure they are also not going to run at startup or from a   scheduled task
    3)Remove Power Manager Driver, then remove Power Manager software
    4)Reboot to Windows
    5)Remove all traces of the Power Manager drivers/software directories (think it was something like C:\readyapps and C:\drivers.) **For some odd reason after I did this, my wireless stopped working but it resolved itself by the time I was done with these steps, strange
    6)Reboot
    7)Access BIOS and reset all settings to default
    8)Boot into Windows, downgrade to BIOS 1.25 [UEFI: 1.25 - 8BET44WW / ECP: 1.14 8AHT32WW ]  via the Windows flash utility. I wanted to downgrade all the way back to 1.06, but the software would throw up some error for any version prior to 1.25 and wouldn’t proceed
    9)Reboot to Windows; make sure system booted w/ out issues
    10)Reboot again, access BIOS, reset to defaults again
    11)Shut down system
    12)Disconnect AC power. Remove main battery. Access and disconnect system backup (a.k.a CMOS) battery under keyboard. Discharge residual power in the system (there are various ways to do this, but you could just leave the battery disconnected for a few minutes).  Visually inspect the system for anything funky…my system had a slightly but noticeably loose CPU/GPU heat sink/fan assembly power connector.
    13)Reconnect backup battery. 
    14)Reconnect AC power but leave main battery disconnected.
    15)Power on. Should get a message indicating “checksum error, system time reset” or something like that.
    16)Reboot to Windows. Verified still okay.
    17)Downloaded latest BIOS version, 1.26. This time I burned the bootable BIOS flash CD instead of running it through Windows.
    18)Restart and boot from disc, flash BIOS to 1.26. Once complete, restart. Verify BIOS set to defaults.
    19)Boot back into Windows.
    20)Install latest Power Manager driver (1.62 ), reboot if/as prompted. Install Power Manager (3.62), reboot as prompted.
    21)Boot into Windows, verified Power Manager was active and working. Verified TurboBoost was working. Restarted system a few times and played around with Power Manager for a bit to see if the different power plans worked and retained the settings, all the while monitoring the Intel TurboBoost utility and PM’s own “power gauges”. Let system Sleep, changed power sources, resumed, etc. Everything was working great.
    22)Shutdown, reconnect main battery. Booted into Windows. Again, fiddled with Power Manager for a bit, switching between power plans and AC/battery power. Still worked great.
    23)Success! 
    That’s it.  Again, this is not a guaranteed fix guide. These are just simply the steps that I took on my system that resolved the problems many of us are having. Hopefully it will work for others.
    T520 4239-CTO | i5 2410M
    W520 4270-CTO | 2720QM | 16GB RAM | Quadro 1000M | BIOS 1.30 | PwrMgr 3.62

    All I basically did was download and install...
    (Chipset driver) http://download.lenovo.com/ibmdl/pub/pc/pccbbs/mobiles/oss924ww.exe
    (PM driver)  http://download.lenovo.com/ibmdl/pub/pc/pccbbs/mobiles/83ku14ww.exe
    Now, I did chipset first (didnt ask to reboot) then installed the PM driver (did ask to reboot)
    I rebooted.... then I went into bios (1.26) and set everything to default... then restarted saving changes...
    Since I prefer not using optimus I changed the display settings in bios right after saving the default settings...
    Not sure whether or not you really had to go into bios... but everything seems to be working...
    My settings in PM is set to Maximum Power in the Advanced tab,  3rd party monitoring tools is TPFanControl and HWInfo64.... 
    W520 (4270 CTO) | i7-2820QM | 16GB RAM 1333 MHz | Runcore MSATA SSD | 2x Kingston SATA2 SSD | Quadro 2000M | FHD | Windows 7

  • Safari CPU scheduling problems

    When I open a tab but remain on the current tab, Safari grind to a halt while the background tab loads. This is NOT a CPU load problem, for I am monitoring the CPU and there is plenty (I am on a G4, so not even clumsy limits to a single core can excuse).
    Things like right-clicking another link to create a new tab takes an age, as does scrolling, which often invokes the beachball. It is as if the foreground page is subordinate to the background tabs.
    This is unacceptable and no excuses outside of bad programming.
    Anyone else see this?
    Message was edited by: Baron Munchausen

    I had the Safari problem in terms of it getting hung up while loading web addresses and had to go through diagnostics every time I moved the laptop from office network to home instead of it switching automatically. How do I tell if Little Snitch is still there? I can see it in the download queue, but no where else. Also, even tho I have to switch manually to the office network and Safari is slow, at least I stay online here without Safari stopping or getting error messages, which makes me wonder if my router is messed up and contributing to problems. I have reset and unplugged my home router with no apparent effect. Sally

  • Qosmio X870 - CPU undervolting

    Continuing my endless search for lower temps, I want to try and undervolt my CPU to get a cooler system.
    I tried using intel XTU but it doesn't seem to let me change anything.
    My processor is the i7 3630QM.
    Is it possible with this processor?
    Anyone know of any ways to mod the BIOS to do it instead?
    Thanks.

    In my opinion such BIOS modification is the wrong way to make the situation better.
    Hardware is tested and some kind of modification is not supported.
    Of course you can do with your machine whatever you want. What is the problem with your machine?
    > Continuing my endless search for lower temps
    I mean if the hardware temperature is normal and acceptable level of working temperature I don't see any reason for such kind of action.

  • CPU Voltage problem...anyone??

    hey all, i'm not sure if i have a voltage problem here.  should my cpu voltage be at 1.5?  it currently runs at 1.35 by default.  This is in my Neo4 Platinum board. 
    Update:
    I went to increased the voltage to 1.4 via CPU VID (in bios) however, CPU-Z is showing my voltage as 1.3 which is lower than before.  is this correct?  Also, watching the voltage readout in CPU-Z i'm noticing the flucuation between 1.36 to 1.408 after setting my CPU VID to 1.4 in bios..  is that ok?
    ...anyone?
    -free

    Just using CoreCenter my 12v is a steady 12.14, 5v at 5.05, 3.3 at 3.33 but vcore which should be the steadiest runs between 1.37 and 1.42 with jumps or drops of up to .04. Prior to wed it never made it past 1.39 so maybe it will even itself out at 1.4  +- .02. My other MSI board a KT6V with a 2500 Barton runs a steady 1.63 with a cheaper PSU and a slight oc and the combined price of the 3 is around what I paid for the neo4 board. Maybe I'm expecting to much to think that if you pay more for something the quality should be better. (this might be old news but look at the class action suit against ABIT for using a defective capacitor on their boards http://www.abitsettlement.com )Quote from: Apple renegade on 18-February-05, 01:25:52
    @home I have a good Volt meter. It cost me more then my mobo. So don't expect the world out the accuricy
    Apple that is like comparing apples and oranges  there is a big difference in a volt meter which has to handle a large range of voltages and amps and a dedicated chip on a motherboard. The CPU is designed so that at 1.4 volt and a given frequency it gives you a signal strength at the transistor that determines a 1 or 0 bit. "undervolt"only increases the chances of bad data if there is a weak transistor and raising the voltage is only needed if the signal strength drops from the increase in frequency.

  • MOVED: MSI Z77A - G45 - CPU multiplier problem

    This topic has been moved to Overclocking, Undervolting.
    https://forum-en.msi.com/index.php?topic=161026.0

    Quote from: xmad on 19-August-12, 07:02:15
    I was just saying that the beta caused a vga problem, latest release the f11 problem so try the release prior to that, but if you got it working don't touch it! lol
    Are you still using the beta bios?
    No clue, we are just users like you are, just keep checking the website. You should report the troubles you had with 2.4 to msi too.>>How to contact MSI.<<
    Yeah iťs working now. As I said, I wanted flash older Bios 2.4, so for sure I click on restore defualts in bios (for sure before flashing) in that 2.64 Bios version...
    PC restarts and my graphic card started working =)...
    So now I am running on that the latest Bios as you refered mi (2.64 beta).
    I am running some test of stability on @ 4 GHz with 1.070 V (IBT,Prime,memmaxx,linx,cinebench,occt,super pi, 3d mark) temps are 62 C max...
    So I guess it's ok now =)
    At the end I have one question :
     My Cpu is now running on 40 x ratio all the time, there is no some underoverlocking ratio (16x etc.) when it's in idle. I was thinking the Ivy Bridge overclocking is overclocking of turbo only...
    I know It's detail, because I have good temps even in idle with 40x ration (33-39 C) but i want to be assured if it is ok.
    EDIT:
    And btw I didnť have 2.4 bios flash on, I wanted put them on... But before flashing my 2.64 beta started working right with HD 7850 as is wrote upper.
    I had Bios 2.5 before 2.64

  • K7T Turbo CPU upgrade problem

    I've got a ms-6330 K7T Turbo limited edition board.  I've been running a 900mhz Duron for years in this board with no problems.  I recently bought an XP1800+ (palomino).  I upgraded to the newest version BIOS, changed the jumper for the 133 FSB and installed the CPU.  The first power-up it hung on the post screen.  After that, I got nothing.  The D-led shows it is hanging at processor initialization.  Put my Duron back in and it works fine.
    I bought the processor off Ebay and realize this may be my problem.  But it was from a reputable seller.  Is there something else I could try before trying to contact the seller for a replacement?  I wish I had another socket A board to try it in, but I don't.

    Are you 100% sure you recieved a Palomino core?
    You can refer to this chart, and compare the CPU to the pictures, to verify that it is a palomino core.
    Also, insure the the heatsink is properly seated, as they will fit the wrong way, but will not make good contact with the CPU.
    Did you use fresh thermal grease on the heatsink?
    PC133 is required to use a 133FSB.
    Try clearing the CMOS, and see if that will allow it to boot.

  • CPU Overheating problem on Windows 8.1 Pro (installed with BootCamp)

    There is a really important overheating problem on windows 8.1 (installed with bootcamp) even i do nothing Cpu is 50-60-70 Celcius degrees. This is really annoying problem. Most of users necessarily need windows OS for using third part programs that doesnt exist on OS X platform. I think this problem is important and has to be fixed as soon as possible.
    My brand new Mac:
    MacBook Pro (Retina, 15-inch, Mid 2014)
    2,5 GHz Intel Core i7
    16 GB 1600 MHz DDR3

    I have the same problem on macbook pro retina 15 late 2013 - top configuration

  • Serious CPU usage problem

    Hi everyone,
    I used to have my own kernel running and everything was fine.
    Recently, after the cool changes which happend with kernel/udev related stuff in the last months and weeks, and with the "beyond" kernel being available and having the possibility to use software suspend2 an easy way,  I thought it would perhaps be better to switch to the beyond kernel package, staying up to date easily and not having to care about wether some update perhaps wouldn't work with my old kernel anymore...
    After some minor hassle, I got it working fine and thought, well, runs reasonably well. But then I discovered some strange issues:
    * If I boot with the regular kernel, log in to the console, use mplayer to play an mp3/ogg file, I got 22% cpu usage by mplayer, and additional system cpu usage, summing up to only 66% of my cpu is in idle state!
    * If I do the same thing with the beyond kernel, it is still around 7 to 8 %, the CPU has a maximum of 87% idle time
    * If I start up X and start xmms, without playing anything, even with stock configuration and nothing changed/enabled, xmms will take 12% of my CPU!
    * If I run some other things aside (like mplayer), xmms - still being idle! - still uses 4% CPU
    * if I actually use xmms, it takes up to 20% cpu
    * now mplayer only uses 2% CPU - but still too much for only playing a music file
    So I just don't have anything left than scratching my head. What's going on there?
    Ford Prefect

    XMMS doesn't play nice with today's Linux scheduler.  It's old.  (I still use XMMS -- out of sheer habit -- and ignore the problem.)   

  • ZIF CPU Upgarde Problems

    I have a B&W Yosemite G3/350,rev 2 board, OS 10.3 & 9.2, the firmware is 1.1.1f4.
    I am trying to change the ZIF with a XLR8 G4/450 and am encountering problems.
    Trying to upgrade the firmware either with Apple's or XLR8 produces -
    "G3 Firmware Updater is for use with the PowerMac G3 only and will not run on this machine" messages,
    this happens if Classic is running or not.
    XlR8's DropFlash application gives "...not compatible hardware..." error.
    Trying to boot with the G4 cpu I get 5 beeps at start up, the same tone as holding the programmers button.
    Grateful for any solutions, keeping the G3 is getting closer to being one.
    Thanks all.
    intel imac 17"   Mac OS X (10.4.5)   PTP225/G4/450/OS9 - powermac G3/OS9&10

    Have you tried an Open Firmware reset?
    To do that, start up, and as soon as you hear the chime, hold down command, option o and f all at the same time until you get a command line.
    Type these in, and hit return at the end of each line.
    reset-NVRAM
    set-defaults
    reset-all
    You might get a blinking "?" for up to a minute, and you might need to reset the time and startup disk.
    "I did a clean install with the OS 9.1 cd with the "update driver" box checked. Since the 9.1 cd would not boot, I installed without doing so."
    Did you do the clean install of OS 9 from "classic mode"?
    If so, I don't know if that would update the driver (classic mode does not allow direct hardware access).
    In OS X, try opening the terminal and using this command
    diskutil list
    This should give you a list of the installed drivers.
    If the OS 9 driver is installed, you should see all of these.
    1: Applepartitionmap 31.5 KB disk0s1
    2: Apple_Driver43 28.0 KB disk0s2
    3: Apple_Driver43 28.0 KB disk0s3
    4: AppleDriverATA 28.0 KB disk0s4
    5: AppleDriverATA 28.0 KB disk0s5
    6: Apple_FWDriver 256.0 KB disk0s6
    7: AppleDriverIOKit 256.0 KB disk0s7
    8: Apple_Patches 256.0 KB disk0s8
    "a copy of OS 9.1 that is on another partition that will not boot either"
    The OS 9 drivers affect the whole drive, not just partitions.

  • CPU utilization problem with JSF

    Hi,
    I am using MyFaces 1.1.4 and tomahawk 1.1.3.
    I am using EJB 2.0 and weblogic server 9.2
    I have a problem with the performance of my application. The response time for 500 concurrent users is arnd 10sec. However, the CPU utilization normally hits *95-97%.*
    I have disabled all the logging to reduce the file IO. Also, all the code has been optimised. Still the problem exists. Now, I cannot think of any way to overcome this. My project is getting delayed because of this.
    Request you all to pls help.
    Regards,
    Milan.

    What type of server or servers are you using. I am assuming which I hate to do is that your EJB's are for CRUD and you a connection pool. Have tried to optimize the DB connection pool's ?.

  • CPU Consumption Problem?

    Hello all,
    I have a 13-inch mb aluminum with snow leopard 10.6 and had a question with a problem that I've been having recently. After doing some searching, I couldn't find anyone with the same problem.
    So basically, what happens is that my computer will be working perfectly fine and the cpu will be idling at 1-3% capacity, which is what it's supposed to do when you're only checking websites/doing word processing ect. Then all of the sudden, my CPU % usage jumps to 50-60% capacity and my laptop starts getting really hot (I have smcfancontrol and it says it's 78C degrees and the fan's speed is 7000rpm). I then proceed to the activity monitor to see what program(s) are sucking the CPU power, but to no avail. All of the processes that are showing aren't taking any more than 4-5% CPU power combined. What gives? This has happened to me twice already in the past 4 days, so I feel that it'll happen again.
    The quick fix is restarting my computer... But I was wondering if there's a more serious problem? Has anybody else experienced this? If so, how did you fix it? I appreciate all of your feedback. Thank you in advance!

    Thank you for your help all! I do have a brother 2170w and did notice the other day that it wasn't working (the wireless part) and had to print elsewhere. So I guess snow leopard is to blame? It all makes sense now, because yeah, I did try to print something after upgrading to snow leopard to no avail.
    Jay,
    I did download and install the new drivers, so hopefully it'll work. I wasn't however, able to find the previous drivers while doing a search in finder (I tried brother, br, and other variations). I just downloaded the new ones and installed it... Think it'll be a problem? Computer's working fine now.

  • CPU Mounting problem

    Hi all i need some help for my m8 as he can not post on the forum.
    ok he has just got a new CPU for his PC now the problem is as he was fitting it he broke one of the clips that fits the CPU to the MOBO and the PC is now over heating due to the CPU not sitting right.
    what i need to no is can you buy this black surround bit from MSI or can he get some sort of cooler that dont use all 4 clips ?
    plz try not and say you need a new MOBO lol but if it is it has to be.
    thanks for any ones help
    Saxon

    There are stores that sell just the Intel Retention Bracket so you can replace the broken one and keep your existing heatsink.  It's like $4 so they are fairly cheap.  I'm having trouble finding the link to the store at the moment.  I will update when I find it.
    edit:  Found the website http://store.yahoo.com/nexfan03/inp4so478rec.html

  • CPU detection problem ?

    I have a 2200+ XP AMD CPU and a KTA4AV mobo.
    Problem is that windows xp says its a AMD 1500 running at 1.35Mhz ??
    Bios is on Auto setting.
    Any idea how I can sort this out ?
    Thx

    Thanks alot, that worked  

  • ITunes high cpu, CoreAUC problem

    Hi,
    Running iTunes 10.4.1 on a 2011 MBA and it's constantly spiking the CPU. I emptied my library, created a new library, tried another account, but the problem won't go away. I don't use any iTunes plugins.
    I've sampled the iTunes process in activity monitor and I see many, many messages like these:
        +                               ! : | + ! : 7 ???  (in CoreAUC)  load address 0x7fff931d9000 + 0x5e6b64  [0x7fff937bfb64]
        +                               ! : | + ! :   2 ???  (in CoreAUC)  load address 0x7fff931d9000 + 0x5e6eef  [0x7fff937bfeef]
        +                               ! : | + ! :   1 ???  (in CoreAUC)  load address 0x7fff931d9000 + 0x5e6e70  [0x7fff937bfe70]
        +                               ! : | + ! :   1 ???  (in CoreAUC)  load address 0x7fff931d9000 + 0x5e6ec2  [0x7fff937bfec2]
        +                               ! : | + ! :   1 ???  (in CoreAUC)  load address 0x7fff931d9000 + 0x5e6edd  [0x7fff937bfedd]
        +                               ! : | + ! :   1 ???  (in CoreAUC)  load address 0x7fff931d9000 + 0x5e6eeb  [0x7fff937bfeeb]
        +                               ! : | + ! :   1 ???  (in CoreAUC)  load address 0x7fff931d9000 + 0x5e6fea  [0x7fff937bffea]
        +                               ! : | + ! 1 ???  (in CoreAUC)  load address 0x7fff931d9000 + 0x59b86e  [0x7fff9377486e]
        +                               ! : | + ! : 1 ???  (in CoreAUC)  load address 0x7fff931d9000 + 0x59c876  [0x7fff93775876]
    Anyone know how I can troubleshoot this problem further ?
    Thanks !

    I am having the EXACT same problem with my MBA 13" mid 2011 ( 128 GB i5 version ).
    I can't find an solution for this and am pretty new to mac OS. Went back to the seller and he said i should format my disk and reinstall everything from a fresh OS copy. Tried it but it didnt fix anything. Went back and he said i should contact apple...
    Also looked on the internet, and the problem isn't that common apparently...
    Found some things about CoreAUC and i'm not sure if replacing it matters, because i already tried formatting it...
    Hope someone can help!
    Thanks in advance

Maybe you are looking for