Overclock.net › Forums › AMD › AMD CPUs › AMD Bulldozer and Piledriver Overclocking Guide - Asus Motherboard
New Posts  All Forums:Forum Nav:

AMD Bulldozer and Piledriver Overclocking Guide - Asus Motherboard - Page 25

Poll Results: Was this helpful?

 
  • 96% (404)
    Yes
  • 3% (13)
    No
417 Total Votes  
post #241 of 4087
Quote:
Originally Posted by aas88keyz View Post



I lost most all stability along with higher core temps. This 3rd time reseating the cpu I finally dropped the temps of core and socket where they used to be. Except then I was at 4.4 GHz and no higher than 55*C on either socket or core. Now I am forcing a 4.3GHz oc and barely getting my cores in at 61*C and socket 55*C. I don't know if I am failing miserably with the TIM or what but I was doing a lot better when I was following your guide in the beginning and then I started messing with things trying to get to a golden overclock (to me atleast) of 4.5GHz. I don't know how much of my TIM I have left and yes I know it only takes the smallest amount but I was just not getting it to spread evenly. I don't know if I should try a reseat again and see how it was spread or accept what I have now. With the temps that I have now do I have the chance to get considerably cooler with reseating once or twice more? I never had this trouble before but I think the difference is the consistency with the TIM. I had more liquidly TIM in the past and now the stuff I have is real thick. though it had a good review. What do you think?

wow my thermalright sb-e extreme is allowing me 4.5ghz with lower temps, yeah arctic silver is good, although I used ceramique ii, quite thick mix, used thermalrights tim that came with the cooler for this build, 3 rice grain sized amounts on cooler plate and same on cpu then used a plastic card (library not bank but same material and thickness) to spread into a thin layer on both cpu and cooler, this makes sure its smooth and evenly spread, then seated the cooler, no need to remove to check spread as you have made sure there is an even coating with the card, hope this helps
post #242 of 4087
Hello,
I'm new in the forum.
I have been trying to OC my FX-8350 for a week and a half now, but can't seem to be able to get any good results with low CPU Voltage
Have tried using only multiplier and also FSB - has not made any difference. 4,6GHz gets stable around 1,45V and 4,7GHz needs ~1,5V.
Stock CPU Voltage is around 1.37V.
I'll add some pictures with my latest settings.


If I decrease Vcore by one point 7th or 8th core will fail in prime within 2 hours.
I hope somebody can help me or will atleast say that my chip is that bad and I should stop trying tongue.gif
DIY
(17 items)
 
  
CPUMotherboardGraphicsRAM
FX-8350 Asus Crosshair V Formula MSI R7850 Power Edition 2GD5/OC CMX8GX3M2A2000C9 
RAMHard DriveHard DriveCooling
CMX8GX3M2A2000C9 INTEL SSDSA2M080G2GC WDC WD10EZRX-00A8LB0 Custom water loop 
OSMonitorKeyboardPower
Windows 8 64bit 24'' Samsung SyncMaster 2493HM Speedlink Bedrock SL6410 Black Saphire Pure PSU 625W  
CaseMouseMouse PadAudio
Wood Logitech MX518  QPAD®|UC™ Large Logitech Z3 
Audio
Roccat Kave 5.1  
  hide details  
Reply
DIY
(17 items)
 
  
CPUMotherboardGraphicsRAM
FX-8350 Asus Crosshair V Formula MSI R7850 Power Edition 2GD5/OC CMX8GX3M2A2000C9 
RAMHard DriveHard DriveCooling
CMX8GX3M2A2000C9 INTEL SSDSA2M080G2GC WDC WD10EZRX-00A8LB0 Custom water loop 
OSMonitorKeyboardPower
Windows 8 64bit 24'' Samsung SyncMaster 2493HM Speedlink Bedrock SL6410 Black Saphire Pure PSU 625W  
CaseMouseMouse PadAudio
Wood Logitech MX518  QPAD®|UC™ Large Logitech Z3 
Audio
Roccat Kave 5.1  
  hide details  
Reply
post #243 of 4087
Dont worry, on my chip i need 1.53v to be stable at 4.7, and around 1.57 for 4.8
post #244 of 4087
Thread Starter 
Quote:
Originally Posted by allavett View Post

Hello,
I'm new in the forum.
I have been trying to OC my FX-8350 for a week and a half now, but can't seem to be able to get any good results with low CPU Voltage
Have tried using only multiplier and also FSB - has not made any difference. 4,6GHz gets stable around 1,45V and 4,7GHz needs ~1,5V.
Stock CPU Voltage is around 1.37V. Warning: Spoiler! (Click to show)
I'll add some pictures with my latest settings.
If I decrease Vcore by one point 7th or 8th core will fail in prime within 2 hours.
I hope somebody can help me or will atleast say that my chip is that bad and I should stop trying tongue.gif

Hopefully I can help you. I see you have a lot of voltages changed and your HTT and CPU/NB are under spec.

1st - Load Optimal Defaults
2nd - See how high you can Overclock just using the CPU Multi on stock voltages using Ultra High LLC.
You can disable all the power saving features and all that, but just leave HTT and CPU/NB on Auto Settings and Auto Voltages.

Test using Prime Small FFT. Just 10 mintues between tests is ok for now.
post #245 of 4087
Quote:
Originally Posted by 1Lt Bob View Post

Dont worry, on my chip i need 1.53v to be stable at 4.7, and around 1.57 for 4.8
Thank you for comforting smile.gif
Quote:
Originally Posted by ComputerRestore View Post

Hopefully I can help you. I see you have a lot of voltages changed and your HTT and CPU/NB are under spec.

1st - Load Optimal Defaults
2nd - See how high you can Overclock just using the CPU Multi on stock voltages using Ultra High LLC.
You can disable all the power saving features and all that, but just leave HTT and CPU/NB on Auto Settings and Auto Voltages.

Test using Prime Small FFT. Just 10 mintues between tests is ok for now.
Ok, did some testing with the following settings:

With stock (1.368750) voltage I was able to run 10 minutes of Prime95 Small FFT up to 4,5GHz. For 4,6GHz I had to up the voltage to 1,393750 (1,404V in CPU-Z). To have 4,7GHz stable in prime small FFT I had to use 1,43750 volts (1,440V in CPU-Z). After running 10 minutes of prime small FFT I tried to run blend, 7th or 8th core failed immediatlly.
Tested same volts with my own settings (previously posted) and got same results. Did not try to get CPU prime95 blend stable takes damn long time. But I have the feeling, it goes the same as it has been before (4,5GHz needs ~1,42V; 4,6GHz needs ~1,46V; 4,7GHz needs ~1,5V).

Will gladly try something new, if you've got some ideas wink.gif
DIY
(17 items)
 
  
CPUMotherboardGraphicsRAM
FX-8350 Asus Crosshair V Formula MSI R7850 Power Edition 2GD5/OC CMX8GX3M2A2000C9 
RAMHard DriveHard DriveCooling
CMX8GX3M2A2000C9 INTEL SSDSA2M080G2GC WDC WD10EZRX-00A8LB0 Custom water loop 
OSMonitorKeyboardPower
Windows 8 64bit 24'' Samsung SyncMaster 2493HM Speedlink Bedrock SL6410 Black Saphire Pure PSU 625W  
CaseMouseMouse PadAudio
Wood Logitech MX518  QPAD®|UC™ Large Logitech Z3 
Audio
Roccat Kave 5.1  
  hide details  
Reply
DIY
(17 items)
 
  
CPUMotherboardGraphicsRAM
FX-8350 Asus Crosshair V Formula MSI R7850 Power Edition 2GD5/OC CMX8GX3M2A2000C9 
RAMHard DriveHard DriveCooling
CMX8GX3M2A2000C9 INTEL SSDSA2M080G2GC WDC WD10EZRX-00A8LB0 Custom water loop 
OSMonitorKeyboardPower
Windows 8 64bit 24'' Samsung SyncMaster 2493HM Speedlink Bedrock SL6410 Black Saphire Pure PSU 625W  
CaseMouseMouse PadAudio
Wood Logitech MX518  QPAD®|UC™ Large Logitech Z3 
Audio
Roccat Kave 5.1  
  hide details  
Reply
post #246 of 4087
Quote:
Originally Posted by amd-pcmarlow View Post

Quote:
Originally Posted by aas88keyz View Post



I lost most all stability along with higher core temps. This 3rd time reseating the cpu I finally dropped the temps of core and socket where they used to be. Except then I was at 4.4 GHz and no higher than 55*C on either socket or core. Now I am forcing a 4.3GHz oc and barely getting my cores in at 61*C and socket 55*C. I don't know if I am failing miserably with the TIM or what but I was doing a lot better when I was following your guide in the beginning and then I started messing with things trying to get to a golden overclock (to me atleast) of 4.5GHz. I don't know how much of my TIM I have left and yes I know it only takes the smallest amount but I was just not getting it to spread evenly. I don't know if I should try a reseat again and see how it was spread or accept what I have now. With the temps that I have now do I have the chance to get considerably cooler with reseating once or twice more? I never had this trouble before but I think the difference is the consistency with the TIM. I had more liquidly TIM in the past and now the stuff I have is real thick. though it had a good review. What do you think?

wow my thermalright sb-e extreme is allowing me 4.5ghz with lower temps, yeah arctic silver is good, although I used ceramique ii, quite thick mix, used thermalrights tim that came with the cooler for this build, 3 rice grain sized amounts on cooler plate and same on cpu then used a plastic card (library not bank but same material and thickness) to spread into a thin layer on both cpu and cooler, this makes sure its smooth and evenly spread, then seated the cooler, no need to remove to check spread as you have made sure there is an even coating with the card, hope this helps

Ok. Duh moment for me... I spaced on the card application. Never done it that way before but I should with the stuff I am using. You totally make sense cause your absolutely right. That is worth a try right away though I will have to pause my boinc crunching from the BOINCers Gone Bonkers event. I will let you all know. I will rep you later.
post #247 of 4087
Quote:
Originally Posted by aas88keyz View Post

Ok. Duh moment for me... I spaced on the card application. Never done it that way before but I should with the stuff I am using. You totally make sense cause your absolutely right. That is worth a try right away though I will have to pause my boinc crunching from the BOINCers Gone Bonkers event. I will let you all know. I will rep you later.

Lol
post #248 of 4087
Thread Starter 
Quote:
Originally Posted by allavett View Post

Thank you for comforting smile.gif
Ok, did some testing with the following settings: Warning: Spoiler! (Click to show)

With stock (1.368750) voltage I was able to run 10 minutes of Prime95 Small FFT up to 4,5GHz. For 4,6GHz I had to up the voltage to 1,393750 (1,404V in CPU-Z). To have 4,7GHz stable in prime small FFT I had to use 1,43750 volts (1,440V in CPU-Z). After running 10 minutes of prime small FFT I tried to run blend, 7th or 8th core failed immediatlly.
Tested same volts with my own settings (previously posted) and got same results. Did not try to get CPU prime95 blend stable takes damn long time. But I have the feeling, it goes the same as it has been before (4,5GHz needs ~1,42V; 4,6GHz needs ~1,46V; 4,7GHz needs ~1,5V).

Will gladly try something new, if you've got some ideas wink.gif

Your voltages running small FFT are right in line with where they should be. You should be able to run Blend at near the same voltages as well.
A couple things to try:
- Make sure you have a current version of Prime95 - 27.7 or newer
- Increase the CPU/NB voltage a little 1.25-1.35v

If Core 7/8 still fail, if it's just a sumout error (xxx was expected, cpu returned xxy) then you can adjust either DRAM VREFDQ or DRAM VREFCA, I can't remember which one, you'll have to do a bit of research. But you can either raise or lower the value so that the CPU returns correct results, without having to increase the cpu voltage.

So basically, you have a very weak 4th module, so you'll have to tweak those settings to keep the voltage down.
post #249 of 4087
Quote:
Originally Posted by ComputerRestore View Post

Your voltages running small FFT are right in line with where they should be. You should be able to run Blend at near the same voltages as well.
A couple things to try:
- Make sure you have a current version of Prime95 - 27.7 or newer
- Increase the CPU/NB voltage a little 1.25-1.35v

If Core 7/8 still fail, if it's just a sumout error (xxx was expected, cpu returned xxy) then you can adjust either DRAM VREFDQ or DRAM VREFCA, I can't remember which one, you'll have to do a bit of research. But you can either raise or lower the value so that the CPU returns correct results, without having to increase the cpu voltage.

So basically, you have a very weak 4th module, so you'll have to tweak those settings to keep the voltage down.
Thanks for these tips will start trying them out right now tongue.gif
DIY
(17 items)
 
  
CPUMotherboardGraphicsRAM
FX-8350 Asus Crosshair V Formula MSI R7850 Power Edition 2GD5/OC CMX8GX3M2A2000C9 
RAMHard DriveHard DriveCooling
CMX8GX3M2A2000C9 INTEL SSDSA2M080G2GC WDC WD10EZRX-00A8LB0 Custom water loop 
OSMonitorKeyboardPower
Windows 8 64bit 24'' Samsung SyncMaster 2493HM Speedlink Bedrock SL6410 Black Saphire Pure PSU 625W  
CaseMouseMouse PadAudio
Wood Logitech MX518  QPAD®|UC™ Large Logitech Z3 
Audio
Roccat Kave 5.1  
  hide details  
Reply
DIY
(17 items)
 
  
CPUMotherboardGraphicsRAM
FX-8350 Asus Crosshair V Formula MSI R7850 Power Edition 2GD5/OC CMX8GX3M2A2000C9 
RAMHard DriveHard DriveCooling
CMX8GX3M2A2000C9 INTEL SSDSA2M080G2GC WDC WD10EZRX-00A8LB0 Custom water loop 
OSMonitorKeyboardPower
Windows 8 64bit 24'' Samsung SyncMaster 2493HM Speedlink Bedrock SL6410 Black Saphire Pure PSU 625W  
CaseMouseMouse PadAudio
Wood Logitech MX518  QPAD®|UC™ Large Logitech Z3 
Audio
Roccat Kave 5.1  
  hide details  
Reply
post #250 of 4087
Ok.. I did the work and I feel satisfied the I was successful at applying the TIM with a credit card. It took me a couple times to find that the kind of card mattered. But that is for another thread. Currently my FX-8120 is 4.2 GHz at stock voltage is even temps between the core and socket at around 55*C. I know I was told that bulldozers need core temps lower than socket temps but am I good when it is even?

Now currently I started Boinc crunching again for the rest of the event. I have to decide when it is over with whether I need to let the TIM cure for a time or if I should start my over clocking over immediately and see if my stability is any better. I have the "80-way Thermal Interface Material Performance Test" link so I can do the research on my TIM whether if it does have a minimum cure time. I can do that on my own. Only thing I don't know is whether I have to be at stock clocks the whole time it takes to cure. I will do what research I can. I just don't know how I was able to get to a cool and stable 4.4 GHz on my 8120 immediately after I finished following this guide and not being satisfied until I could make 4.5 GHz I lost my footage and barely make a hot but stable 4.2 GHz since then. Any advice is welcome.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD CPUs
Overclock.net › Forums › AMD › AMD CPUs › AMD Bulldozer and Piledriver Overclocking Guide - Asus Motherboard