Forum Home
Press F1
Thread ID: 110863	2010-07-04 22:52:00	why OCing screwed up windows?	powerover (12121)	Press F1

Post ID	Timestamp	Content	User
1115849	2010-07-04 22:52:00	Hi there guys, I OCed a bit too far the other week, try to go for the 4GHz zone and my mobo won't boot, not even post. Ended up resetting the CMOS jumper to clear the BIOS setting. now here is the thing, with all factory settings windows won't boot properly any more, it freezes half way through the boot up animation (when the four blobs of color come together to form the windows logo), then blue screen of death. I ended up reinstalling windows. Before the OC everything was fine, no trouble at all. I am using windows 7 Pro 64bit. I did try using the start up repair thing windows offered, but no, doesn't work at all. My question is, if my mobo didn't even post, that means it didn't even start windows right? how did windows screw up??? which part of windows screwed up? thanks for the help guys.	powerover (12121)
1115850	2010-07-04 23:10:00	You obviously havn't had much experience with overclocking, I have corrupted windows on at least a dozen occasions from overclocking too far, I even killed a CPU once! I can't tell you why it happens, I can only guess that it is when the disk is reading during startup, it writes back files to the disk (probably event logging etc) and that is when corruption occurs.	Deimos (5715)
1115851	2010-07-04 23:39:00	You obviously havn't had much experience with overclocking, I have corrupted windows on at least a dozen occasions from overclocking too far, I even killed a CPU once! I can't tell you why it happens, I can only guess that it is when the disk is reading during startup, it writes back files to the disk (probably event logging etc) and that is when corruption occurs. Nice...you must be a guru then, prepare to be bombarded with OCing questions :P right now my CPU seems to be stable at 3.825GHz, it passed the OCCT 1hour test, and a OCCT 2.5hours test. when I was doing the first test Vcore was set at 1.3325V, then it passed the test, then I fire up youtube and watched a video, during which graphics driver crashed and recovered. then Blue screen of death...:(...then I updated the driver to the latest. then I thought Vcore is a bit too high so set it at 1 notch down from 1.3325V (cnt remember the exact number, but 1 minimum step down my mobo allows), then it passed the second OCCT test (the 2.5 hours one). then I put my pc to sleep, will play with it again tonight. my ram voltage is set at 2.2, with 2.1 it wnt even post with my mobo..:(, rams are running at 851MHz now it is a long shot, but can you tell what caused the blue screen of death after passing the first 1hour test? is it the ram voltage too low? or my power supply not juicy enough? or anyother factor?? thank for the help :D	powerover (12121)
1115852	2010-07-05 09:06:00	There is no way I can tell you what is causing your instability, you just have to fiddle with it on your own to find out (sometimes it can take weeks to find a stable OC). But first of all, as far as I'm aware, no overclocking vet uses OCCT, the only tool I trust with a quick and dirty OC test is Intel Burn test (akak IBT, good for C2D or C2Q) or Linx (for C2D, C2Q, i3, i5 and i7) both of these programs are actually Linpack with a different front end but IBT doesn't seem to stress the new i3, i5, i7s very well) Just an FYI, if you have pushed your CPU too hard too soon, IBT will kill your CPU (thats how I killed my first CPU), then for long term stability testing, the latest version of Prime95 (multi threaded support) for 24 - 48 hours. You must do at least 4 passes of IBT or Linx for a quick and dirty test (this should be your first test when testing a new OC, I go with 8-10 passes to be sure, but you MUST do a minimum of 4) and it is brutal, in can find instability when Prime95 goes for 24hours without errors, and you will notice temps of 10C higher (or more if you are on air) than any other load you can possibly throw at it, for example, while gaming my CPU and Vidcard both hit the mid to high 40s (water) but with LinX the CPU alone hits 77! Once you have found an OC that you are happy with, then you run 20+ (I wouldn't go over 30) runs of IBT or LinX (10 passes is all it took to kill my C2D E6600 @ 3.6Ghz on water) followed by 24+ hours of Prime95	Deimos (5715)
1115853	2010-07-05 09:09:00	Oh yeah, and your PSU looks a little weak to be overclocking with SLI, I tend to go way overboard with mine, you can see my specs here: New PC (pressf1.pcworld.co.nz)	Deimos (5715)
1115854	2010-07-05 13:06:00	OP hasnt said anything about cooling or temps. And I havent OC'd an 8400 before but those volts seem a little high?	hueybot3000 (3646)
1115855	2010-07-05 13:58:00	On some motherboards, overclocking the CPU also overclocks the SATA bus. Any hard drives connected will then get problems with data corruption. It can also happen that overclocked CPU\RAM\Chipset is beyond its limits and is corrupting data before it gets written to the drive.	Agent_24 (57)
1115856	2010-07-05 22:13:00	is it the ram voltage too low? or my power supply not juicy enough? or anyother factor?? AFAIK, that RAM voltage should be 2.2 for 800-1066 speed.. Here is the best clocking guide on the planet. Thats Anandtech forums.anandtech.com	SolMiester (139)
1115857	2010-07-05 23:21:00	Oh yeah, and your PSU looks a little weak to be overclocking with SLI, I tend to go way overboard with mine, you can see my specs here: New PC ( . pcworld . co . nz/showthread . php?t=110802" target="_blank">pressf1 . pcworld . co . nz) umm . . good point, I am probably pushing it a bit . but I used this "online extreme power supply calculator" thing, with OCing considered a factor, it tells me that I only need 543W . . . . And I use to use prime95, it takes aaaages, and I saw somewhere online that someone was saying that OCCT does it in 1hour = prime95 does it in 24hours so I switched . Guess patients is the key huh?? :P Thanks for the help and advice . :D	powerover (12121)
1115858	2010-07-05 23:25:00	OP hasnt said anything about cooling or temps. And I havent OC'd an 8400 before but those volts seem a little high? thanks for the reply:D I use to OC the exact same CPU and ram on a SLI ready mobo, using the same voltage (they read the same under CPUz) and it ran very stable, never crashed once. I guess different chip sets likes it differently? what I discover here, from my personal experience, is that SLI chipset mobos are much easier to OC than the intel chipset mobos, because in SLI chipset mobos you can unlink the FSB and the ram frequency, but not in intel chipsets...at least the mobo im using now anyway..:badpc:	powerover (12121)
1 2 3