• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

WTF is wrong with my computer?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

SMOKEU

Member
Joined
Nov 7, 2010
Location
NZ
So, I built this computer up 2 years ago. Since then, I've had numerous motherboard failures, and the computer has never been quite "right".

Whenever I overclock the CPU, I run Prime95 for 24+ consecutive hours, and do the same with Intel Burn Test, except for 12+ hours until I have no stability issues. After a few months of rock solid performance, the system starts giving all sorts of stability issues that nothing short of an OS reinstall seems to fix. So far I've had to reinstall the OS about 2-3 times a year at least due to this problem.

I started a thread on here near the end of last year to get help overclocking my i7 to be stable at 4.7GHz. It performed flawlessly with over 100 hours of stress testing all together, and I had no problems with it for months. Then, the GPU drivers started giving me severe problems with artifacting, flickering etc so I uninstalled the old drivers, used CCleaner and DriverSweeper to remove the traces of the old drivers, and after testing 6 different driver versions I gave up and reinstalled the OS (I set the CPU clocks to standard before the OS reinstall like I always do).

So recently I decided to load my old 4.7GHz stable BIOS settings (I saved the OC profile in the BIOS), and it gives nothing but problems. It runs 1.400V on average at that speed and while it's generally stable for most 12+ hour Prime95 runs, Intel Burn Test usually crashes after about a minute or 2, as does OCCT. I reduced the clock speed to 4.6GHz, and played around with the offset voltage so it's about 1.375V, and I still have the same problems.

I'm starting to get really sick of this temperamental build that only wants to work properly for a few months before this happens. I have just done a 8+ hour Memtest 86+ 4.20 run with no errors. It doesn't seem to give any problems at stock settings. If I had enough money I would shoot this computer or throw it into the sea and build a new one, but unfortunately I'm flat broke and won't be able to afford anything better than a dual core Pentium with onboard graphics if I do that.

When gaming the screen sometimes go blank, but the monitor light stays blue which means it's still getting some sort of signal (otherwise it turns orange-yellow colour). The event logs say the "system rebooted from a bugcheck" or "an I/O operation initiated by the Registry failed unrecoverably. The registry could not flush the hive (file):". Temperatures are not the problem since it still keeps crashing even on a cold morning when the hottest core is under 70°C.

I have included some screenshots of the BIOS to show the once stable 4.7GHz settings that are no longer stable.

IMG_20130604_174630.jpg

IMG_20130604_174545.jpg

IMG_20130604_174609.jpg

I have updated the BIOS a few months ago to the latest version.
 
Last edited:
Two settings I'd change right off the bat... VCCIO voltage to 1.15V and CPU Current Capability to 110-120%. And instead of a fixed VRM frequency of 350, try setting it to Auto.
 
Two settings I'd change right off the bat... VCCIO voltage to 1.15V and CPU Current Capability to 110-120%. And instead of a fixed VRM frequency of 350, try setting it to Auto.

I tried that, and it passed Intel Burn Test for the default 10 runs, then I tried again and it crashed within a couple of minutes.
 
I set the VRM frequency back to 350 and enabled PLL overvoltage, as well as CPU current capability to 120%, and it passed 10 runs of IBT, then I tried again and it failed on the first run. I don't understand why it's sometimes stable and sometimes not. It seems to have a mind all of its own.
 
Error about failing to flush the registry changes, reeks of bus instability or RAM instability.
 
Error about failing to flush the registry changes, reeks of bus instability or RAM instability.

Hence the reason why I ran Memtest. I also ran Prime95 on blend mode which is meant to test a lot of RAM, and didn't have any problems with 12+ hour runs on Prime95 or Memtest.

Intel Burn Test also sometimes gives errors at stock BIOS settings (I reset everything back to stock, except for manually entering the standard RAM speed, timings and voltage as printed on the RAM labels).

I'm really after more of a long term solution here as I know for a fact that if I "fix" it somehow, it's only going to buy me a few months at most before it's deja vu all over again and I'll be back here starting another thread like this.

It's never been right since new, but I could never take it back for a warranty claim and say "it sometimes works fine, but sometimes doesn't".
 
Hmm. Wonder if the OP has a really really good surge protector or a UPS for his PC. He could have major power issues that keep frying his stuff. He could have a really bad ground or lots of things.

If the OP has replaced xx stuff, time to replace something else.

Or run it at stock for 6 months. Maybe run memtest for 5 days. 4.6 isn't a massive OC you could have a long term stability issue tho depending on your CPU.

Temps good?
 
Hmm. Wonder if the OP has a really really good surge protector or a UPS for his PC. He could have major power issues that keep frying his stuff. He could have a really bad ground or lots of things.

If the OP has replaced xx stuff, time to replace something else.

Or run it at stock for 6 months. Maybe run memtest for 5 days. 4.6 isn't a massive OC you could have a long term stability issue tho depending on your CPU.

Temps good?

I have got a surge protector connected to this computer (I've already replaced it). I used to have the computer connected to a UPS, but then the UPS blew up before sending 275V into my computer (I've had these problems long before I had a UPS, so it's not the UPS that stuffed the computer up).

Since it's winter the CPU stays below 75°C on the hottest core. At the moment it seems to be stable at 4.6GHz for OCCT and Prime95 for 8+ hour runs, but it still crashes randomly and gives a BSOD or the screen just goes blank.
 
whats the actual vcore when overclocked whilst running p95 and IBT??? did you have it set to monitor whilst running those??? Most agree that anything over 1.40 for a period of time starts the early degradation of the cpu. Cold or not. Most have seen there once rock hard cpu overclock loose about 100-200Mhz and get stable again only to falter again later for another 100-200Mhz loss til even running stock without raising vcore wont stay stable.
 
whats the actual vcore when overclocked whilst running p95 and IBT??? did you have it set to monitor whilst running those??? Most agree that anything over 1.40 for a period of time starts the early degradation of the cpu. Cold or not. Most have seen there once rock hard cpu overclock loose about 100-200Mhz and get stable again only to falter again later for another 100-200Mhz loss til even running stock without raising vcore wont stay stable.

With IBT for the short amount of time that it runs, the vcore is around 1.400V, while Prime95 is 1.376-1.39V. OCCT also stays around 1.39-1.400V.
 
I have got a surge protector connected to this computer (I've already replaced it). I used to have the computer connected to a UPS, but then the UPS blew up before sending 275V into my computer (I've had these problems long before I had a UPS, so it's not the UPS that stuffed the computer up).

Since it's winter the CPU stays below 75°C on the hottest core. At the moment it seems to be stable at 4.6GHz for OCCT and Prime95 for 8+ hour runs, but it still crashes randomly and gives a BSOD or the screen just goes blank.

Drop the OC to 4.0. Try it. 6 months.
 
So 2 days ago I got it stable at 4.6GHz and it passed a few 10+ hour runs of OCCT and Prime95. Now it's not even stable at stock clocks.

If I can't even get it stable at stock clocks, what's the next step to take?
 
sounds alot like the cpu has degraded if you have a spare to test it out and confirm its not the mobo having issues keeping volts flowing. Or vice versa if you have a spare mobo it'll confirm the same findings.
 
sounds alot like the cpu has degraded if you have a spare to test it out and confirm its not the mobo having issues keeping volts flowing. Or vice versa if you have a spare mobo it'll confirm the same findings.

I don't have a spare CPU and I don't know anyone who does. I might buy a cheap Celeron dual core like a G1610 since they're the cheapest LGA1155 chip I can find. My board should work with any LGA1155 chip with the latest BIOS wouldn't it?

If I buy the Celeron at least I will will get 2/3 of my money back when I sell it.
 
Possibly faulty caps on the board out-of-the-box! :mad:

Gotten reports of this happening before, which is as rare as a hibiscus blooming in Alaska!
 
Back