FX 4170 When overclocked, unstable

[clay30x]

Forgot about that, rebooted and changed to 1.25 was running at 1.09 or so on Auto.

 

[clay30x]

Going to read the guide, Thanks.
I appreciate your patience. For me it's been a pain.

 

[Johan45]

I've gotta take off clay. Will try and pop in later if you do go playing around just monitor your temps use prime95 for testing and keeep it below 70 on the cpu and 60 on the core.

 

[RGone]

How to put up Sig at OCF.

New Shortcut method for putting a Signature with your system information following your every post so people can know what is in the case that they are trying to assist with. You can use something like what is shown in my signature as a good template of needed information Thank you.

 

[clay30x]

Thank you for that, will update asap.

Btw, Johan, that Guide helped tremendously. I did a moderate overclock of the CPU, I am stable at 4.69 Ghz. I had no idea about the 2 Northbridges on AMD Bulldozer..
I don't want to go too high at the moment, without monitoring my temps for a bit. I'm still in the break in period for my arctic silver.

I'm thinking about posting a few screenshots of the ASUS UEFI Bios for overclockers that have the Sabertooth.. Might help, make my own guide.

Thanks again

 

[RGone]

Sabertooth 990FX should have CPU_NB listed and is the only one we really worry about both for voltage to it and the speed. Default speed is about 2200Mhz and to raise it any will likely take about 1.25V on the CPU_NB.

HT Link speed is 2600Mhz by default however some bioses set that to 2200Mhz as well, this was especially so on the first bios version or so. Raising the HT Link speed requires no additional voltage adjustments or 'never' has for me since it is the actual default speed AMD set forth. I never up the other NB voltage.

F-12 in the bios will save the exact portion of the bios that you are seeing on the screen to a BMP file on a Usb 2.0 stick formatted to FAT 32. The file sizes are big. 2MB or so and I open in graphic editor and save as JPG at about 170K so a good bit smaller file size. It takes me a couple of F-12 saves to actually save the entire page of say the Voltage settings. LLC page can be saved in one F-12 and so it goes thru the bios.
RGone...

 

[Johan45]

Thank you for that, will update asap.

Btw, Johan, that Guide helped tremendously. I did a moderate overclock of the CPU, I am stable at 4.69 Ghz. I had no idea about the 2 Northbridges on AMD Bulldozer..
I don't want to go too high at the moment, without monitoring my temps for a bit. I'm still in the break in period for my arctic silver.

I'm thinking about posting a few screenshots of the ASUS UEFI Bios for overclockers that have the Sabertooth.. Might help, make my own guide.

Thanks again

Way to go clay! A good 2 hour session of P95 will let you know if you're stable or not and help with the arctic burn in!! Just pay attention to the temps.

 

[clay30x]

RGONE, would having my CPU_NB at 2500+mhz and set to. 1.27Ghz cause a crash, I'll let you know of my symptom that started last night:


So I started crashing last night after a reboot attempt on my machine.
I'm starting to think that it's something to do with the chip, possibly a lemon or an issue related in this post, although old:
http://scalibq.wordpress.com/2011/10/19/amd-bulldozer-can-it-get-even-worse/

This is the symptom:
- I rebooted after installing software.
- After windows login, looking at my Windows Gadgets, I notice one of the Cores (CORE 1) sticks at 100% while my system loads, the CPU obviously being heavily utilized. It never goes down to below 100% and 5-15 seconds later, blue screen.
NOTE: This is not a new problem. This has happened before but I didn't think it was a seperate issue until now, becuase I think my voltages are fine.

- Thinking it was a voltage problem, I uped the voltage on my CPU and CPU_NB by .01 CPU 1.38 and CPU_NB 1.28
After reboot, it still crashed but it crashed when CORE 2 went to 100% for about 10 seconds.

- I set it back to previous settings becuase prior, it was running for almost 24hrs or so at 4.7GHZ. Rebooted, and still same thing happened with another core (Core 1)

- Skip to this morning. I booted the system and it started stable. No crash. I did see your post and prior to this set the CPU_NB frequency to 2200. It was set to 2500Mhz.
The HT Frequency is 2500mhz and change.

I'll keep it like this for a time and see what it brings me. I am now running BOINC on it so I'll see if anything should happen today while that runs.

 

[Johan45]

RGONE, would having my CPU_NB at 2500+mhz and set to. 1.27Ghz cause a crash, I'll let you know of my symptom that started last night:


So I started crashing last night after a reboot attempt on my machine.
I'm starting to think that it's something to do with the chip, possibly a lemon or an issue related in this post, although old:
http://scalibq.wordpress.com/2011/10/19/amd-bulldozer-can-it-get-even-worse/

This is the symptom:
- I rebooted after installing software.
- After windows login, looking at my Windows Gadgets, I notice one of the Cores (CORE 1) sticks at 100% while my system loads, the CPU obviously being heavily utilized. It never goes down to below 100% and 5-15 seconds later, blue screen.
NOTE: This is not a new problem. This has happened before but I didn't think it was a seperate issue until now, becuase I think my voltages are fine.

- Thinking it was a voltage problem, I uped the voltage on my CPU and CPU_NB by .01 CPU 1.38 and CPU_NB 1.28
After reboot, it still crashed but it crashed when CORE 2 went to 100% for about 10 seconds.

- I set it back to previous settings becuase prior, it was running for almost 24hrs or so at 4.7GHZ. Rebooted, and still same thing happened with another core (Core 1)

- Skip to this morning. I booted the system and it started stable. No crash. I did see your post and prior to this set the CPU_NB frequency to 2200. It was set to 2500Mhz.
The HT Frequency is 2500mhz and change.

I'll keep it like this for a time and see what it brings me. I am now running BOINC on it so I'll see if anything should happen today while that runs.

A few things occur to me right off the hop. First have you tested it for stability with P95? If you were able to run consistently and then crash after a software install, I'd lean more in that direction. Software conflicts or even malware. When is the last time you ran a scan?

If you haven't done a Prime run for a couple hours then yes you could be unstable at those clocks be it CPU clock or NB clock.

 

[clay30x]

P95: I have for about 20 minutes.

The software was BOINC, and the core locked up before BOINC even ran. I have experienced this before BOINC was installed. It's usually after the system is on for a long time and I do a reboot for some reason.

This symptom that happens is rare, but when it happnes, it happens consistently. (maybe irrelevant.)

The last time I ran a Malware scan was very recent and actually should be a non issue because I reformatted a month back. But regardless. I will do 2 things.

1. I will run Malware Bytes. (A "Full" scan could take some time as I have about 2.2 TB on the system.

2. I will run P95 for a few hours at a different time.

(BTW, I have never crashed the system while running P95 on this system while overclocked. I've had errors on the cores but only that. (Right now p95 runs without error. I'll schedule a time, tomorrow to run it.

Thanks!

 

[Johan45]

I would swap out that psu too. Just to eliminate it. You say that it's troublesome after hours of use which could be heat related and that's your weakest like from what I can see.

 

[clay30x]

Did a Full Scan with Malware Bytes and Windows Defender. Both had 1 or 2 items on my storage drive but just pirated keygens and stuff. Nothing on C:\ where Windows is.

After downclocking the CPU_NB to 2200mhz and keeping the CPU_NB voltage at 1.27 which is 1.269 I think, I haven't had a crash.
After a good 3 or 4 hours of no crash, I decided to switch out my PSU back to the Seasonic which is just a better quality PSU. Seasonic SS-620GM (620W)

I'll probably upgrade to a better PSU soon after I get some money allocated.