Jump to content

X_X

Member
  • Posts

    426
  • Joined

  • Last visited

Reputation Activity

  1. Informative
    X_X got a reaction from thekingofmonks in How to adjust memory timings on a laptop.   
    Download RWEverything from here. Be careful how you use this program as exception handling is poor meaning BSOD is easily achieved with wrong input.
     
    Click first button, "PCI Devices" and select "SMBUS Controller"

     
    Double click on the dword at Reg 0x40 and check bit 4 is set to 0.

     
    If it's not zero then SMBUS SPD writes are disabled so either a BIOS mod is required or the SODIMM needs to be programmed externally, example here
     
    Some SPD chips have protection from writing to the first 128 bytes, if it's hard set then the SPD chip will need to be changed out, if soft set then usually a special voltage needs to be set on one of the SPD chip's SMBUS address select pins. Need to check relevant SPD chip data sheet.
     
    Other option is to program an XMP profile which comes after the first 128 bytes but XMP will need to be selected in the BIOS.

    If external programming then any mistakes can easily be rectified by reprogramming and of course before doing any programming a backup of the SPD data should be made for each module by simply reading the data into a file.
     
    If programming in SITU then only use one SODIMM first. If it's bad it can be replaced with the unmodified SO-DIMM to POST and then usually the bad SO-DIMM can be inserted as it is usual for the BIOS to remove power to the empty slot but you need to check to be certain this is the case. The Bad SO-DIMM SPD can still be accessed and reprogrammed.
     
    Find out how to do a CMOS reset and check that it works before you change anything.
    RWEverything can also be used to program SPD data.

    FWIW Personally for my Core 2 Duo laptop I programmed the SPD simply because it was hardware modified to run BCLK at 333MHz instead of 266MHz. The RAM needed to be adjusted for the higher clock or it would not post. By doing it through SPD there wasn't any problem if a reset occurred. That laptop did many hours and still works today. My Haswell laptop had an option to enable hidden settings without a modified BIOS although I modified the BIOS anyway. At the time I bought it, it had only one DDR3L SO-DIMM and DDR3L was hard to find locally at the time so ended up with 2 SO-DIMMS from different manufacturers but timing was much the same. I used the BIOS settings to increase from 1600MT/s to 1867MT/s and it's been working fine like that for the last 5 years.




     
    If this is your BIOS then these are the options it contains including the hidden ones. IFR.zip
     
    You might find this thread interesting but whatever you do always have a recovery plan and part of that means taking backups before you start modifying. You can get some good advice as to what to do by posting on Win-raid.com.
     
  2. Informative
    X_X got a reaction from Firebreath335 in 3200mhz ram clocked at 1600mhz?   
    Your DRAM uses a 1600MHz clock. DDR double data rate means it can transfer data on both rising and falling edges of the clock meaning it can transfer 2 bits of data per clock. IOW 3200 Mega Transfers per second or 3200MT/s.
    Understand it for what it is. If all manufacturers stuck to advertising the real clocks then this confusion would not come about and you would not be here asking your question.
     
    FYI the current world record for DDR4 (which is liquid nitrogen cooled) is 2783MHz so don't go thinking your RAM is running at 3200MHz. Once one manufacturer advertises twice the clock frequency (as effective frequency compared to non DDR from a long time ago) then the others have to follow or lose sales because the average person will assume the double frequency is better.
     
    Similar has happened with disk drives, instead of the binary kilobyte (1024) they use the metric kilo (1000) so a 186GiB drive shows up as a 200GB drive. So far RAM sizes are safe as a 16GB DIMM is actually a little over 17GB using the metric measurement.
     
    Another way to look at it is when say a 4GHz CPU operates in 32-bit mode it uses 32-bit general purpose registers but in 64-bit mode it uses 64-bit general purpose registers which are double the width and can transfer twice the data per clock but it doesn't mean it's operating at 8GHz.
     
    TLDR No, it's not slow.
  3. Informative
    X_X got a reaction from pc_user in Turn off red audio boost, LED light. MSI Gaming 5   
    Settings->Advanced->Power Management Setup->On board Function LED Control
  4. Informative
    X_X got a reaction from ujLion in How to adjust memory timings on a laptop.   
    Download RWEverything from here. Be careful how you use this program as exception handling is poor meaning BSOD is easily achieved with wrong input.
     
    Click first button, "PCI Devices" and select "SMBUS Controller"

     
    Double click on the dword at Reg 0x40 and check bit 4 is set to 0.

     
    If it's not zero then SMBUS SPD writes are disabled so either a BIOS mod is required or the SODIMM needs to be programmed externally, example here
     
    Some SPD chips have protection from writing to the first 128 bytes, if it's hard set then the SPD chip will need to be changed out, if soft set then usually a special voltage needs to be set on one of the SPD chip's SMBUS address select pins. Need to check relevant SPD chip data sheet.
     
    Other option is to program an XMP profile which comes after the first 128 bytes but XMP will need to be selected in the BIOS.

    If external programming then any mistakes can easily be rectified by reprogramming and of course before doing any programming a backup of the SPD data should be made for each module by simply reading the data into a file.
     
    If programming in SITU then only use one SODIMM first. If it's bad it can be replaced with the unmodified SO-DIMM to POST and then usually the bad SO-DIMM can be inserted as it is usual for the BIOS to remove power to the empty slot but you need to check to be certain this is the case. The Bad SO-DIMM SPD can still be accessed and reprogrammed.
     
    Find out how to do a CMOS reset and check that it works before you change anything.
    RWEverything can also be used to program SPD data.

    FWIW Personally for my Core 2 Duo laptop I programmed the SPD simply because it was hardware modified to run BCLK at 333MHz instead of 266MHz. The RAM needed to be adjusted for the higher clock or it would not post. By doing it through SPD there wasn't any problem if a reset occurred. That laptop did many hours and still works today. My Haswell laptop had an option to enable hidden settings without a modified BIOS although I modified the BIOS anyway. At the time I bought it, it had only one DDR3L SO-DIMM and DDR3L was hard to find locally at the time so ended up with 2 SO-DIMMS from different manufacturers but timing was much the same. I used the BIOS settings to increase from 1600MT/s to 1867MT/s and it's been working fine like that for the last 5 years.




     
    If this is your BIOS then these are the options it contains including the hidden ones. IFR.zip
     
    You might find this thread interesting but whatever you do always have a recovery plan and part of that means taking backups before you start modifying. You can get some good advice as to what to do by posting on Win-raid.com.
     
  5. Like
    X_X got a reaction from Bitter in Ghetto Intel Core i7 from CHINA for $100!!   
    Runs at a fixed 1600Mz which is tied to BCLK so while you cannot adjust ratio's AFAIK you can use the 125MHz strap for 2000MHz.
  6. Agree
    X_X got a reaction from skycat2216 in How intel or AMD lock they're CPU multiplier   
    In that one you can check that P_State_Limits isn't limiting your ratio but a lot of the settings you want are int the SDM mostly under the MSR section such as MSR 0x198, 0x199, 0x1A0. Limits may or may not exist in both MSR and MMIO so you would need to check especially as some might not be documented.
  7. Informative
    X_X got a reaction from TsanakJim in How to adjust memory timings on a laptop.   
    Download RWEverything from here. Be careful how you use this program as exception handling is poor meaning BSOD is easily achieved with wrong input.
     
    Click first button, "PCI Devices" and select "SMBUS Controller"

     
    Double click on the dword at Reg 0x40 and check bit 4 is set to 0.

     
    If it's not zero then SMBUS SPD writes are disabled so either a BIOS mod is required or the SODIMM needs to be programmed externally, example here
     
    Some SPD chips have protection from writing to the first 128 bytes, if it's hard set then the SPD chip will need to be changed out, if soft set then usually a special voltage needs to be set on one of the SPD chip's SMBUS address select pins. Need to check relevant SPD chip data sheet.
     
    Other option is to program an XMP profile which comes after the first 128 bytes but XMP will need to be selected in the BIOS.

    If external programming then any mistakes can easily be rectified by reprogramming and of course before doing any programming a backup of the SPD data should be made for each module by simply reading the data into a file.
     
    If programming in SITU then only use one SODIMM first. If it's bad it can be replaced with the unmodified SO-DIMM to POST and then usually the bad SO-DIMM can be inserted as it is usual for the BIOS to remove power to the empty slot but you need to check to be certain this is the case. The Bad SO-DIMM SPD can still be accessed and reprogrammed.
     
    Find out how to do a CMOS reset and check that it works before you change anything.
    RWEverything can also be used to program SPD data.

    FWIW Personally for my Core 2 Duo laptop I programmed the SPD simply because it was hardware modified to run BCLK at 333MHz instead of 266MHz. The RAM needed to be adjusted for the higher clock or it would not post. By doing it through SPD there wasn't any problem if a reset occurred. That laptop did many hours and still works today. My Haswell laptop had an option to enable hidden settings without a modified BIOS although I modified the BIOS anyway. At the time I bought it, it had only one DDR3L SO-DIMM and DDR3L was hard to find locally at the time so ended up with 2 SO-DIMMS from different manufacturers but timing was much the same. I used the BIOS settings to increase from 1600MT/s to 1867MT/s and it's been working fine like that for the last 5 years.




     
    If this is your BIOS then these are the options it contains including the hidden ones. IFR.zip
     
    You might find this thread interesting but whatever you do always have a recovery plan and part of that means taking backups before you start modifying. You can get some good advice as to what to do by posting on Win-raid.com.
     
  8. Informative
    X_X got a reaction from KhakiHat in Is it worth the effort to overclock?   
    Be warned that it ca be addictive :)
     
    Nothing wrong with software overclocking if done properly.
  9. Like
    X_X got a reaction from aezakmi in Dead AIO-cooler, or did I overlook something?   
    I had a H100i stop cooling recently after about 3 years use and google seems to show the same thing time and again, blockages. There were also some stories of the warranty replacements not lasting long either so I decided to void my warranty (5 years) and open it up.
     
    Liquid port obstructed

     
     
    Channel for copper cooling fins blocked.

     
    It looks to me like something has crystallized in the liquid, perhaps some silicone grease that wasn't right for the job. I would think Corsair would be very aware of the problem after receiving returns and have hopefully addressed the issue in their newer units.
     
    At some time I'll try to put it back together minus the blockages and see how it goes. Pump was still turning but maybe starved of flow or dead heading into the blockage. After all it was fine when it was working properly so maybe worth a shot.
     
    Possibly in the Op's case there was some build up in the cooling circuit and moving the unit around dislodged it into a blockage. Just a guess.
     
  10. Like
    X_X got a reaction from TheBean in cinebench r20   
    Hwinfo CPU Tab example

     
    For undervolting there is a guide, software and help here
     
  11. Informative
    X_X got a reaction from Spotty in CPU Throttling   
    If the heatsink on the power stage isn't making good thermal contact it might not be obvious that things are getting hot.
     
    If you have warranty that might be a good way to go unless they give you back something crappy
     
    HWiNFO supports the ISL6388 on your board, perhaps it will show the actual TM temperature and at what value VR_HOT# is activated. The temperature may also be used as part of the compensation circuit so better if it's working properly. Your original BIOS also has an option under "Digital Power" to set "CPU VRM Over Temperature Protection" from 85C - 120C, default is Auto. Actually it has an option to disable it too but it's hidden from the user.
     
    Check see if those HWiNFO temperature registers are shown, might also have the power in/out values too, and you can use these possibly as ammo for your warranty claim.
  12. Like
    X_X reacted to Bitter in Ghetto Intel Core i7 from CHINA for $100!!   
    Well I'm seeing core temps of under 70C, so I can't imagine it's a thermal issue but maybe it's more aggressive since it's a laptop chip. The 4980HQ in her MacBook Pro thermal throttles at like 99C pretty hardcore because MacBook Pro.
     
    As long as it's stable I won't worry too much, I need to swap cooler and get it installed into the HTPC case along with the better PSU and RAM that's not over clocked so I'm probably going to see less stability issues with better parts. I'll work on messing with the microcode in Win10 once it's installed, no sense doing it on the test right with 7 since that's not the HTPC disk.
  13. Like
    X_X got a reaction from Bitter in Ghetto Intel Core i7 from CHINA for $100!!   
    Voltage settings are 1 to 1. IIRC resolution of VID is 4-5mV, not analogue and it also can change with temperature.
     
    Microcode is temporary, IOW when the CPU is reset it loses any microcode update and needs to be reloaded so with the VMWare solution it would only last until you uninstalled the VMWare driver. Once uninstalled you would be back to your original setup.
     
    Validation looks fine.
  14. Like
    X_X got a reaction from Bitter in Ghetto Intel Core i7 from CHINA for $100!!   
    Perhaps or because the change wasn't enough to hit the next step or it dropped because of temperature rise. My 4700MQ VID would change by 50mV over a 60C delta when using dynamic voltage. As temperature increases the VID decreases.
  15. Funny
    X_X got a reaction from Jumper118 in Post your Cinebench R20+15+R11.5+2003 Scores **Don't Read The OP PLZ**   
    Looks good to me but it's @Jumper118 you need to convince. Better hope someone hasn't stolen the jam from his donut.
     
    Not sure there's a system slow enough and meets the minimum requirements of CBR20 to score an MT of 1 cb if run at defaults.
     
     
  16. Like
    X_X got a reaction from TheBean in cinebench r20   
    Here's one for i7-4980HQ, took a lot more than 45W though !

     
    For the 8750H there is one posted on HWBOT just 19pts shy of 3000 cb.
     
  17. Like
    X_X got a reaction from PianoPlayer88Key in Post your Cinebench R20+15+R11.5+2003 Scores **Don't Read The OP PLZ**   
    Looks good to me but it's @Jumper118 you need to convince. Better hope someone hasn't stolen the jam from his donut.
     
    Not sure there's a system slow enough and meets the minimum requirements of CBR20 to score an MT of 1 cb if run at defaults.
     
     
  18. Like
    X_X got a reaction from Mr. Question Asker in Should I update my BIOS?   
    Okay  ?
  19. Like
    X_X got a reaction from Mr. Question Asker in Should I update my BIOS?   
    ??? Yes, I am aware of the performance losses which is why I said it conflicts with the BIOS description in the Op's first post.
  20. Informative
    X_X got a reaction from Mr. Question Asker in Should I update my BIOS?   
    Kind of conflicts with "2. Improved system performance and stability"
     
    Would have thought Spectre fix would have been addressed earlier so maybe this is a better fix that mitigates some of that performance loss. With such an outright ambiguous description who knows.
     
    OP, check your BIOS settings and make a copy you can refer to, make a backup of your current BIOS and if rolling back is supported I would suggest giving it a go. I guess your running Gigabyte with dual BIOS so should be okay, if your comfortable doing it.
  21. Like
    X_X reacted to Jurrunio in Adaptive mode   
    yes they still happen. Removing voltage spikes completely requires predicting the future accurately and consistently which we cannot do atm.
     
    No, manual is no worse than adaptive
     
    Use adaptive for vcore to reduce power draw at low loads.
     
    Voltage spike isn't much of a problem on Haswell and other architectures with integrated voltage regulator. By putting the voltage regulator inside the CPU, it can react to voltage spikes much faster by changing duty cycles of the high side fet earlier (shorter distance for sensing signals to move around). If you're really concerned you can increase the FIVR switching frequency, but note that this comes at the cost of efficiency and heat output.
  22. Agree
    X_X reacted to dgsddfgdfhgs in Cpu advice? Please?   
    never trust any saleman. they got commision for what they do.
    9700k does NOT have HT/ smt  , typically its not  I7.. 2700 is much better in editing
  23. Like
    X_X reacted to Tech Wizard in I undervolted my GTX 1080 to 0.7V - Results   
    That undervolt was under load. When idling it's sitting at 139MHz / 0.6V
  24. Like
    X_X got a reaction from Tech Wizard in I undervolted my GTX 1080 to 0.7V - Results   
    @Tech Wizard Curious, if trying to save power then why the fixed voltage and clock, why not let them drop down when idle?
     
    FWIW my 1050Ti does 1400 GPU clock at 0.7V (0.84V using multimeter) and crashes at 1500 so your doing a whole lot better than me
     
    At 1400 Heaven shows 68W using GPU-z but 50W using nvsmi and Hwinfo. I think GPU-z is calculating it using the TDP % and default power which leads to incorrect results when default power is changed - bug. Not sure it really matters in this case as nvidia has said 1050Ti power readings are crap, so much so that they removed them from the latest drivers.
     

     
  25. Informative
    X_X got a reaction from Hiro Hamada in TDP for 1070Ti   
    Not necessarily, for Intel the actual hardware shutdown CPU temperature is around 125C-130C. Now your software ACPI may be set to shutdown before then otherwise providing thermal monitoring isn't disabled or power management hasn't been initialized such as Skylake non-k OC then with zero offset applied to the Temperature Target aka Tjmax the processor will throttle at 100C in an attempt to keep the temperature from increasing.
     
    Similar with nVidia GPU, you can set a cap for temperature which when exceeded will throttle down to base clock and only appears to further throttle when near maximum settable temperature. For my 1050Ti that is 97C, when hitting 96C it will throttle below base clock otherwise only down clocks to base clock. There are also hardware limits of 99C for slowdown temperature whatever that means exactly and shutdown temp is 102C. For my 1080 it's 96C and 99C respectively.
     
    You can check with the nvsmi utility that usually installs with the nvidia drivers. "nvidia-smi.exe -q -d temperature" 
     
    It's nice to have these features should the cooling fan(s) stop for some reason such as a bug or failure or water pump for that matter.
×