AX850 going bad or something else?

I am having an intermittent BIOS boot failure during cold/warm boots or restarts. It happens randomly (i.e. not every time). Basically, on some boots/restarts, the computer will try to POST and then apparently fail and give me the "BIOS boot failure" message followed by options to load optimized defaults or enter the BIOS to reconfigure settings. All I do is enter the BIOS (which will be in a default state (i.e. with XMP disabled and everything else at default settings) and hit "save and exit" because the settings are retained but just not active. Once the computer is on and in Windows, everything is fine and I can do anything and there are no issues including long gaming sessions. It's only (randomly) during cold/warm boots or restarts.

 

One thing I noticed is that this only started occurring right after I overclocked my 3930K from it's stock 3.8 GHz to 4.0 GHz. I didn't change anything other than the CPU multiplier. It was a very mild overclock.

 

Here are the settings prior to the issue:

 

3930K at 3.8 GHz: 1.200 volts

DRAM (DDR3-2133 CAS 9): 1.5 volts

IMC: 1.000 volts

VTT: 1.05 volts

 

Here are the settings after the overclock and when the issue started happening:

 

3930K at 4 GHz: 1.200 volts

DRAM (DDR3-2133 CAS 9): 1.5 volts

IMC: 1.000 volts

VTT: 1.05 volts

 

Things I've tried to fix the issue that did not help:

 

1. Increase Vcore

2. Increase IMC voltage to 1.05 volts

3. Increase DRAM voltage to 1.525 volts

4. Replace the CMOS battery

5. Remove the overclock and go back to the settings prior to having this issue

6. Run MemTest86 for 4 hours which reported no errors

 

I searched on the Internet and it seems issues like this could point to the power supply or motherboard. Any help will be appreciated.

 

By the way, I built this computer in January, 2012 so that should give you an idea of the age of most of the equipment (other than the graphics card).

 

Edit: I also do not have any spare computer parts available to test with my computer.

Run everything stock for a while n see if it still happens. Probably not psu issue.

By "stock" do you mean XMP enabled or disabled? I have already tried going back to the settings I was using prior to having this issue (no overclock and XMP enabled) and still encountered this issue.

 

Can bad RAM cause something like this?

Overclocked my 3930K to 4.2 GHz. I started at 1.200 Vcore and ran IntelBurnTest at the maximum stress level. Kept increasing the Vcore till the BSODs stopped. They eventually stopped when I set a Vcore of 1.230 volts. Let's see how this turns out.

Just happened again. I've noticed one thing though. During a successful POST boot, during the POST the fans in my case, including the H100's fan's, run at a very slightly lower speed and once the POST is completed successfully they run at their full required speed. During to a boot/restart that results in a BIOS boot failure, the fans would be running at an even lower speed as if the computer is not getting enough power. Basically, I've noticed that boots/restarts that result in a BIOS boot failure message all the fans in my case would be running at a lower speed than normal as if the computer is not getting enough power. Bad power supply or motherboard?

All fans, excluding the H100's fans, are connected to my case's (650D) fan controller which, if I remember correctly, is connected to the power supply with a molex connector. The H100's fans are connected directly to the motherboard running at full speed all the time.

 

Note that out of all the BIOS boot failure messages I've gotten, there have been 2-3 times when the computer failed to POST and instead of getting the BIOS boot failure message, the computer just shut down.

 

Edit: With my 4.2 GHz overclock on my 3930K with a Vcore of 1.230 volts, I ran 10 instances of IntelBurnTest at the maximum stress level. At the end of the test (after the 10 runs were completed), I got a message saying my overclock(?) was not stable and that I should reduce my overclock. However, I do not encounter any BSODs when running IntelBurnTest with a Vcore of 1.230 volts but do with anything lower. When I run just 1 instance of IntelBurnTest, at the end of the test I get a "successful run" or the like message. Do I need to increase my Vcore ever so slightly?

Slow down only during a POST but ever so slightly. It's always been like this, even when the computer was new. Not sure if my case fans also slow down (the ones connected to the fan controller) because the H100's fans are the ones I hear and the rest of the case fans are pretty quiet and hard to hear.

 

Note that this only started happening after I overclocked my 3930K from its stock 3.8 GHz to 4.0 GHz. Once it started happening, I keep increasing the Vcore but it made no difference. Now, even if I return to no overclock and reset all settings to the way they were before the overclock, I still get this issue. Lastly, it's only during a (random) POST and once I am in Windows there are no issues of any sort. Maybe I'll just live with it. It's been happening since December, 2015.

Would damaged CPU VRs on the motherboard only cause issues with POSTs? Usage within Windows is 100% fine.

Prior to having this issue, I was running at 3.8 GHz at 1.200 Vcore. When I overclocked my processor to 4.0 GHz at the same Vcore (1.200) I started having this issue. I thought it was because I needed more Vcore. I slowly kept upping the Vcore till 1.215 but it kept happening. Could a 200 MHz overclock without raising the Vcore damage the CPU VRs on the motherboard? Right now I'm running 4.2 GHz at 1.230 Vcore.

Sounds like bios glitch to me more then anything. Like when your rig is off your getting a cmos clear or something. If it runs fine once in Windows there's no issue with anything hardware level I don't think. Try doing a cmos clear and if that doesn't work update to latest bios.

Just happened today again during a restart after installing the latest NVIDIA drivers. I am already on the latest BIOS and there are no more releases for my motherboard (old X79 chipset). Should I try re-flashing it?

Just happened today again during a restart after installing the latest NVIDIA drivers. I am already on the latest BIOS and there are no more releases for my motherboard (old X79 chipset). Should I try re-flashing it?

 

Try clear cmos.

Isn't that the same as loading optimized defaults? I don't have a CMOS reset button. I have to short it using a jumper or similar. I also recently replaced the CMOS battery. I removed the battery, waited about a minute or so, then put in the new battery. Wouldn't this count as resetting the CMOS?

Yes but did you replace battery because of this issue or did you replace it and then the issue started ?

Replaced it because of this issue.

Next logical step for me would be so an overnight run of memtest86. Other then that I'm gonna have to do some research on this one. I remember an issue like this from a few years back I'll see if I can find anything on it.

http://sanuja.com/blog/how-to-fix-overclocking-failed-error

This is not about gigabyte board but the Intel settings in the 2nd part of the fix may work for you. Disabling virtualization tech etc. Also there is a recall on your motherboard if you didn't know. Google gigabyte x79-ud3 recall. Might get an upgraded motherboard out of it lol.

Thanks.

 

Happened today again but this time during POST the computer shut down. Left the computer on during the night and when I restarted it this morning, the computer shut down during POST. When I turned it back on, I got the BIOS boot failure message that I always get when this happens. Seems all this is more likely during a warm boot/restart instead of the cold boot/restart.

 

I will try to loosen RAM timings or reduce the RAM clock speed. Will also run MemTest86. By the way, should I run MemTest86 with all my RAM in the computer or run it with one stick at a time (I have four sticks of 4 GB)?

Ran MemTest86 overnight with XMP enabled. Completed 4 passes in 6 hours and 30 mins with no reported errors. As soon as I restarted the computer, I encountered the BIOS boot failure message. I even tried disabling XMP and running my RAM at 1333 MHz and still encountered the BIOS boot failure message during a restart. I guess it's not the RAM. Right now I have changed my LLC setting in the BIOS from Auto to Medium. Let's see if it helps.

 

Also found this thread with a similar issue: https://forums.anandtech.com/threads/2500k-occasional-post-failure.2443336/

Do you have c states enabled and speedstep etc. All the power saving features. Also c1e disabled of course ?