2500k CPU voltage stability?

[Ika]

I have a Gigabyte Z68X-UD3H board with my i5 2500k. I'm working on overclocking now and while I'm happy with what I have, I'm trying to see if I can get the temp levels (ergo, noise levels) down a little bit. Right now I'm shooting for a decent 4.5ghz OC.

With the Gigabyte "Multi-Steps Load Line" set to level 6 (what appears to be the most level vdroop control) and BIOS voltage set to 1.295, I'm 18h Prime95 stable and at least one hour Linpack stable (still running). However, when I tried 1.280v, I got an immediate BSOD upon loading Windows. Okay, reasonable. 1.285v. BSOD. 1.290v. BSOD again. 1.295. Totally stable.

So, what gives? Is the 2500k just that picky about its voltages? Should I be concerned about running the voltage so close to unstable BSOD voltage? My temps are hitting around 65-68C max running Prime95 and Linpack with the motherboard out in open-air (my case fans haven't arrived yet) at the 1.295v level, with a Hyper212+ and stock fan.

 

[Puppies04]

How high are you trying to overclock?

 

[Ika]

How high are you trying to overclock?

Whoops - I have it at 45x multi (100.2 base clock, mobo shows it around 100.1 or so).

 

[Ika]

Well, turns out I wasn't really fully stable at all. At 1.295v in BIOS, 1.31/1.32 in TouchBIOS, when I booted into Windows Prime95 was more likely stable than not - had 3h, 5h, 8h, and 17h fully stable. I couldn't always boot into windows, though, BSODing at random times. I thought it was my memory or something.

I bumped vcore to 1.310v in BIOS and it's reading 1.32/1.33 in TouchBIOS. So far no BSODs. What a strange chip.