Lesser known dangers of Overclocking

Associate
Joined
20 Sep 2006
Posts
1,418
Location
Southampton, England
Thought i'd have a fun eve pushing my system to the limits, and it was rather educational in a number of ways...... some good, some bad.

Firstly, i can get my E6400 boot stable at 3.6Ghz/1.425V (didnt do any stability testing), at which point it benches about the same on Sandra as a Q6300 (1.86Ghz quad core) for ALU/FPU.

Secondly, it IS possible to corrupt your windows instalation through overclocking (tried to push up to 3.7Ghz). Thankfully i had some unformated space on my new HDD, and all my important docs are in a seperate partition on the new HDD, so i just formated a new partition, installed windows, and can pull any important stuff off the "busted windows" bit (like god knows how many emails in Thunderbird).

Anyone else had anything like that? not just comp reboots and BSODs
 
Instability can cause data corruption yes, thats 1 reason why its a good idea to use a proper stress test.
 
You shouldnt really install the OS OC'ed unless you know you have a working OC, if thats what you're saying.
 
My raid setup got borked when i overclocked once,had to do a fresh install to fix it,my single drive seem`s to hold out better for overclocking.
 
Windows makes several disk writes during startup, if the system is 'very' unstable yes, it can write data to the wrong locations on the hard drive, potentially overwriting, and corrupting the windows installation.

I have a 'test' CD which loads Memtest 86, and when pushing the limits of an overclock I run this before I boot up my windows. If it cant pass 1 complete cycle of memtest, then its quite likely to corrupt the windows install.

If there are no errors after 5 passes, then even though the system may still fail orthos, at least its not totally unstable, and naturally that reduces the risk of disk corruption.

As already mentioned it 'can' also be due to running the PCI/SATA/PCIe buses out of spec,
 
Back
Top Bottom