CPU the problem here?

[Nafrayu]

A week ago I experienced some issues with my PC, and have been dealing with them since. I posted another thread here about checking to find if I had issues with my HDs, both of which came back 100% healthy in crystaldisk. Among my issues were frequent wow crashes, reoccuring system lockups, and BSODs

I then went and cleaned my pc, format and all, reinstalled windows and began to have issues with windows not wanting to update, hanging, etc

Since then I have spent about 10 hours with microsoft support attempting to get my windows to update, every time they attempt something it does nothing to fix the issues, or it causes a BSOD.

Originally built in march I OC'd upto 4.5 @1.3 v and found it was completely stable under a prime 95 overnight test.
During my issues I went into BIOS and changed the loadout back to default settings, so I could have as little complication as possible.

Today I decided to place my OC back to what it was(was going to run a stability test in PRIME, see if I would get bluescreens after nuking the drive) and strangely could never get my PC to boot. What once were stable numbers, now do not work at all. I also noticed that after one of the crashes, bios settings had changed my multiplier to 48 instead of 45, and that my voltage got changed back to Auto.

Im really confused at this point, and any help would be greatly appreciated.

[chadashcroft22]

Ran a memtest? My guess is bad stick(s) of ram. And wouldn't hurt to run sfc/scannow to make sure no windows files are corrupt: http://pcsupport.about.com/od/toolso...fc-scannow.htm

[Tradewind]

What is the BSOD stop/error code you are getting?

Sounds similar to my problem recently with my 2600K, causing 0x101 BSOD's. Ended up being the CPU.

[Deleted]

really make sure that all bios settings are the same as they were compared to the prior overclock.
faulty ram and mobo could also be a cause of problems.

[c0rnel]

I believe its your overclock. Instead of 4.5 try 4.0 at 1.3v. Not all chips are the same. I had the same issue with my cpu at 4.5. It just needs more volts but the cpu really starts to get hot evenwith watercooling

[fatalwario]

go back to stock settings, check mem, etc

[Nafrayu]

really make sure that all bios settings are the same as they were compared to the prior overclock.
faulty ram and mobo could also be a cause of problems.

I have gone to my best lengths to ensure every setting is back to what I had changed it to. XMP, DDR changed to 1600mhz, Manual for the overclock, multiplier 45, PLL overvolt off, and voltage set at 1.3. I also went about trying other overclocks, such as 3.5-4.0 with varying voltages of 1.2-1.3 and nothing was stable enough to load windows. It would get partway into a load, and crash, either blue screening like I mention, or hardlocking like I mention.

What is the BSOD stop/error code you are getting?

Sounds similar to my problem recently with my 2600K, causing 0x101 BSOD's. Ended up being the CPU.

I put my OC back to 4.5, with the correct settings, and the blue screens either cause a hard lockup, or the screen flashes blue for what seems like a millisecond before the computer restarts itself. in the event viewer the common errors codes are 0xC00000D4. hr=0x80070057 and this one 0x0000001e (0xffffffffc0000005, 0xfffff803966e2cb3, 0x0000000000000001, 0x0000000000000003). A dump was saved in: C:\Windows\MEMORY.DMP. Report Id: 070213-2375-01. Several warnings over A corrected hardware error has occurred.

Reported by component: Processor Core
Error Source: Corrected Machine Check
Error Type: Internal parity error
Processor APIC ID: 2

Ran a memtest? My guess is bad stick(s) of ram. And wouldn't hurt to run sfc/scannow to make sure no windows files are corrupt: http://pcsupport.about.com/od/toolso...fc-scannow.htm

I ran memtest and let it complete, it found no errors.
I ran the SFC scan 3 times total. First time it failed at 47%. Second time it passed. Ran it a third time to make sure and a blue screen occured(same thing happened days ago while Microsoft support ran the same scan)

[chadashcroft22]

Check all your temps and voltages on everything in your system with HWMonitor. We will want to know cpu temps with no load and fully loaded. And the most important reading for voltage is what your 12v rail reads loaded and without load. http://www.cpuid.com/softwares/hwmonitor.html

[Tradewind]

Right click on Computer in the start menu, then click on Properties, on the left side go to Advanced system settings, then the Advanced tab, click Settings under Startup and Recovery. Under System failure, uncheck Automatically Restart and change the debugging information to Small memory dump.

This will stop the computer from automatically rebooting when the BSOD shows up, giving you time to write down the Stop code and any associated driver causing the problem.

http://www.nirsoft.net/utils/blue_screen_view.html

is also a helpful tool for examining minidump's after a bsod to find faults and error codes/parameters.

[Nafrayu]

Check all your temps and voltages on everything in your system with HWMonitor. We will want to know cpu temps with no load and fully loaded. And the most important reading for voltage is what your 12v rail reads loaded and without load. http://www.cpuid.com/softwares/hwmonitor.html

I got an image of the temps on a fresh boot, and again 15 minutes later after some web browsing and following TradewindNQ's advice, however when attempting to run prime95 it caused a BSOD under each of the 3 different settings it has(small large and blend). I found a simple browser based one but it didnt seem to be very effective, though it did cause my temperatures to rise to just about 70.

No load http://i44.tinypic.com/24pz3w0.png
15 mins of web browsing and event viewinghttp://i44.tinypic.com/2nm0pbk.png
load from webpage after 15 minshttp://i41.tinypic.com/xbxt2w.png

Right click on Computer in the start menu, then click on Properties, on the left side go to Advanced system settings, then the Advanced tab, click Settings under Startup and Recovery. Under System failure, uncheck Automatically Restart and change the debugging information to Small memory dump.

This will stop the computer from automatically rebooting when the BSOD shows up, giving you time to write down the Stop code and any associated driver causing the problem.

http://www.nirsoft.net/utils/blue_screen_view.html

is also a helpful tool for examining minidump's after a bsod to find faults and error codes/parameters.

okay I grabbed that program since Im on Win8 and couldnt quite figure out where to find out how to follow your instructions on it. The returned dumps are there, but Im not overly sure which information from it is important. one says page fault in nonpaged area, another clock wathdog timeout and another kmode exception not handled


One thing I noticed while running the webbased CPU stress test, was that the temperatures on my cores fluctuated heavily. Not only across cores as the image shows, but it would also quickly jump from 30-60 back to 39, back to 50ish, back down to 30's. And not just one core, but seemingly all of them.

[Tradewind]

"clock watchdog timeout"

that's the one I'm actually concerned with, 0x101. If you set everything back to stock and run a Prime 95 test (stress only > Blend), does it BSOD almost instantly?

Honestly, if you can get a hold of another Socket 1155 processor, try it. Doesn't have to be anything fancy, if you can even just borrow one for a day or two.

[Nafrayu]

"clock watchdog timeout"

that's the one I'm actually concerned with, 0x101. If you set everything back to stock and run a Prime 95 test (stress only > Blend), does it BSOD almost instantly?

Honestly, if you can get a hold of another Socket 1155 processor, try it. Doesn't have to be anything fancy, if you can even just borrow one for a day or two.

Next to instantly. Either the program and my PC hangup for a few moments before a bluescreen, or I barely have enough time to react before it blue screens. And unfortunately, aside from buying another processor, I don't have access to another socket 1155 processor.

That code in particular was the one leading me to believe it was CPU related, as many support sites I had read said that has to do with your Clock speeds.

[Tradewind]

For me it was my 2600k and somehow being incapable of regulating voltages anymore. I hate suggesting it because it's kind of a dick thing to do lol, but maybe just find somewhere that has a reasonable return policy and buy a cheap i3 or something :P Somewhere you could return it, if it's not the actual problem. Honestly, testing to find out will take all of an hour...you'll be able to tell with prime 95 after swapping the CPU.

But your problem sounds identical to my own issue just a couple weeks ago:

http://www.mmo-champion.com/threads/...n-BSOD-Lockups

and it necessitated a new CPU.

[Nafrayu]

For me it was my 2600k and somehow being incapable of regulating voltages anymore. I hate suggesting it because it's kind of a dick thing to do lol, but maybe just find somewhere that has a reasonable return policy and buy a cheap i3 or something :P Somewhere you could return it, if it's not the actual problem. Honestly, testing to find out will take all of an hour...you'll be able to tell with prime 95 after swapping the CPU.

But your problem sounds identical to my own issue just a couple weeks ago:

http://www.mmo-champion.com/threads/...n-BSOD-Lockups

and it necessitated a new CPU.

Blah, definitely not what I wanted to hear, but what I was dreading after the research on clockwatchdog lol. I'm going to have to look into that in the area(only a very few limited retailers in the area)

Thank you a million for the help though. And from reading through your post, the problems seem identical(aside from you not having a MS update issue)

[chadashcroft22]

Vcore should NEVER be more then 1.38v period most sandybridge/ivy bridge chips dont need more then 1.20 even for beyond 4Ghz.

Seems as though you've done some damage running such a high vcore for so long. I hate to say it but that was the risk you take when you start pumping up voltages in cpus. Odds are there is nothing you can do beyond replacing the cpu and I would recommend not overclocking above stock voltages --no matter what enthusiasts tell you-- You can get to 4ghz on stock voltages with most anything up untill haswell. Any gain after that isn't worth the extra heat/lower lifespan of your system.

[Faithh]

Board you are using? This doesn't sound like a cpu issue and 1.3V is just nothing. Try setting the VCCSA voltage to 1.10V and cpu pll voltage to 1.80V. Most of the time it's just the board refusing an overclock. You might just want to try your ram in another system and maybe if you like the cpu as well but I'm kinda certain your board is dying.

- - - Updated - - -

Vcore should NEVER be more then 1.38v period most sandybridge/ivy bridge chips dont need more then 1.20 even for beyond 4Ghz.

Seems as though you've done some damage running such a high vcore for so long. I hate to say it but that was the risk you take when you start pumping up voltages in cpus. Odds are there is nothing you can do beyond replacing the cpu and I would recommend not overclocking above stock voltages --no matter what enthusiasts tell you-- You can get to 4ghz on stock voltages with most anything up untill haswell. Any gain after that isn't worth the extra heat/lower lifespan of your system.

If you load an xmp profile, the clock speed is for a 2600k 3800MHz even if 4 cores are being used so the extra 200MHz (4GHz oc) you gain isn't worth it compared to 4.5GHz overclock.

[Nafrayu]

Board you are using? This doesn't sound like a cpu issue and 1.3V is just nothing. Try setting the VCCSA voltage to 1.10V and cpu pll voltage to 1.80V. Most of the time it's just the board refusing an overclock. You might just want to try your ram in another system and maybe if you like the cpu as well but I'm kinda certain your board is dying.

The board I am using is an ASUS p8z77-v LK. I would try the ram and CPU in another system, if I actually had another to try it on >_> Everyone and their mother that I know uses prebuilt craptops.

I tried changing the settings as you said, but was unable to get the CPU PLL voltage to change, it has a setting of Auto, or +0.1(at least as far as I could see)