- Nov 17, 2005
- 86
- 0
- 0
Simply put, can anything go wrong with a computer when overclocking besides losing data on the hard drive or overheating the cpu and gpu?
Originally posted by: scrawnypaleguy
What order would you recommend doing overclocking? Up the FSB until it won't post, and then increase voltage? What about ram timings/frequency? Also, what exactly happens when an attempted overclock "doesn't work"? Your computer refuses to boot... then what? How do you get it back to stock so that you can change things?
Originally posted by: Matt2
Originally posted by: scrawnypaleguy
What order would you recommend doing overclocking? Up the FSB until it won't post, and then increase voltage? What about ram timings/frequency? Also, what exactly happens when an attempted overclock "doesn't work"? Your computer refuses to boot... then what? How do you get it back to stock so that you can change things?
If you go to high the comp will not boot or post. at that point you're going to have to reset the CMOS to get back to defaults, then up the voltage from there.
Originally posted by: cpacini
Originally posted by: Matt2
Originally posted by: scrawnypaleguy
What order would you recommend doing overclocking? Up the FSB until it won't post, and then increase voltage? What about ram timings/frequency? Also, what exactly happens when an attempted overclock "doesn't work"? Your computer refuses to boot... then what? How do you get it back to stock so that you can change things?
If you go to high the comp will not boot or post. at that point you're going to have to reset the CMOS to get back to defaults, then up the voltage from there.
Depending on your board, you can also hold down a key on the keyboard (usualy insert) when booting, and it will boot the machine in a "safe" mode at default settings.
Originally posted by: Acanthus
Originally posted by: cpacini
Originally posted by: Matt2
Originally posted by: scrawnypaleguy
What order would you recommend doing overclocking? Up the FSB until it won't post, and then increase voltage? What about ram timings/frequency? Also, what exactly happens when an attempted overclock "doesn't work"? Your computer refuses to boot... then what? How do you get it back to stock so that you can change things?
If you go to high the comp will not boot or post. at that point you're going to have to reset the CMOS to get back to defaults, then up the voltage from there.
Depending on your board, you can also hold down a key on the keyboard (usualy insert) when booting, and it will boot the machine in a "safe" mode at default settings.
DFI boards can also save up to 5 bios configurations for fast switching, helps a ton :thumbsup:
Originally posted by: biostud
I think I might have damaged my CPU due to o/c. Even running at stock speed now I sometimes get errors in files when downloading stuff. (My maximum was o/c 1.55V 2.7Ghz, not stable)
The frequency of errors increase as I o/c the CPU more. It could be that the o/c had damaged some windows files, but I doubt that's the problem.
So if I'm right in my conclusion o/c can damage the CPU permanently, but if anyone can prove me wrong I would be glad.
Originally posted by: Acanthus
Originally posted by: biostud
I think I might have damaged my CPU due to o/c. Even running at stock speed now I sometimes get errors in files when downloading stuff. (My maximum was o/c 1.55V 2.7Ghz, not stable)
The frequency of errors increase as I o/c the CPU more. It could be that the o/c had damaged some windows files, but I doubt that's the problem.
So if I'm right in my conclusion o/c can damage the CPU permanently, but if anyone can prove me wrong I would be glad.
That is almost ALWAYS the case with overclocking too far.
Id reformat.
Originally posted by: biostud
Originally posted by: Acanthus
Originally posted by: biostud
I think I might have damaged my CPU due to o/c. Even running at stock speed now I sometimes get errors in files when downloading stuff. (My maximum was o/c 1.55V 2.7Ghz, not stable)
The frequency of errors increase as I o/c the CPU more. It could be that the o/c had damaged some windows files, but I doubt that's the problem.
So if I'm right in my conclusion o/c can damage the CPU permanently, but if anyone can prove me wrong I would be glad.
That is almost ALWAYS the case with overclocking too far.
Id reformat.
That would be really sweet if that's the case
Originally posted by: GuitarDaddy
Originally posted by: biostud
Originally posted by: Acanthus
Originally posted by: biostud
I think I might have damaged my CPU due to o/c. Even running at stock speed now I sometimes get errors in files when downloading stuff. (My maximum was o/c 1.55V 2.7Ghz, not stable)
The frequency of errors increase as I o/c the CPU more. It could be that the o/c had damaged some windows files, but I doubt that's the problem.
So if I'm right in my conclusion o/c can damage the CPU permanently, but if anyone can prove me wrong I would be glad.
That is almost ALWAYS the case with overclocking too far.
Id reformat.
That would be really sweet if that's the case
Almost certainly the case! A CPU usually works or it doesn't, if you fry it its dead, I've never seen one that sporatically causes errors due to damage from OC. What you describe is a corrupted windows install, should be fixed by a reinstall or a repair install of window. If the boot sector has been corrupted then if may require a full format and reinstall
Originally posted by: biostud
I think I might have damaged my CPU due to o/c. Even running at stock speed now I sometimes get errors in files when downloading stuff. (My maximum was o/c 1.55V 2.7Ghz, not stable)
The frequency of errors increase as I o/c the CPU more. It could be that the o/c had damaged some windows files, but I doubt that's the problem.
So if I'm right in my conclusion o/c can damage the CPU permanently, but if anyone can prove me wrong I would be glad.
That is not correct as there are could be part of the execution unit or one or more of the execution units that's damaged so in OOO operation you could still get something that's ok but when the schedular goes out and assign the bad one you could be in troulbe. It could also depend on the interconnect width, weather it was level 1, 2 or 3 interconnect and the thickness of the via to connect the layers. So you could get a partially working chip.Originally posted by: GuitarDaddy
Originally posted by: biostud
Originally posted by: Acanthus
Originally posted by: biostud
I think I might have damaged my CPU due to o/c. Even running at stock speed now I sometimes get errors in files when downloading stuff. (My maximum was o/c 1.55V 2.7Ghz, not stable)
The frequency of errors increase as I o/c the CPU more. It could be that the o/c had damaged some windows files, but I doubt that's the problem.
So if I'm right in my conclusion o/c can damage the CPU permanently, but if anyone can prove me wrong I would be glad.
That is almost ALWAYS the case with overclocking too far.
Id reformat.
That would be really sweet if that's the case
Almost certainly the case! A CPU usually works or it doesn't, if you fry it its dead, I've never seen one that sporatically causes errors due to damage from OC. What you describe is a corrupted windows install, should be fixed by a reinstall or a repair install of window. If the boot sector has been corrupted then if may require a full format and reinstall
Originally posted by: Chesebert
very possible and propable as the interconnect or via could be damaged due to electromigration with such high power output that they are never designed to handle nevermind the temperature. you are upping the volts to 1.55 which is 15% over stock and the poewr usage would probably be 40-50% greater becase of the v^2 relationship and the high leakage current for SOI. Your interconnect and vias would fail more quickly than the actual gates as they are littl more resiliant.
Originally posted by: biostud
Originally posted by: Chesebert
very possible and propable as the interconnect or via could be damaged due to electromigration with such high power output that they are never designed to handle nevermind the temperature. you are upping the volts to 1.55 which is 15% over stock and the poewr usage would probably be 40-50% greater becase of the v^2 relationship and the high leakage current for SOI. Your interconnect and vias would fail more quickly than the actual gates as they are littl more resiliant.
the temperature was not that high (I'm on water), but it could ofcourse be higher on the actual die. Currently the re-install seems to have fixed the problem. I've downloaded lots of drivers etc, with no errors so far, so hopefully it was just a software error.
One thing is sure though 1.45v 2.6Ghz is absolute maximum for o/c this chip I'm willing to run.