Hey guys, I've been able to successfully OC my 2600k to 4.8 GHz, with 1.35 Vcc, but if I run Prime95, it will start failing in less than one minute. On the other hand, setting the Vcc to 1.45v will make it run without errors but that value's proximity to the 1.52 "Official" limit is worrying me. What exactly am I doing wrong? My temps are mostly fine with the new A50 cooler, but I'm totally lost. What's worse is that others can Prime95 totally stable on 4.8 with 1.35.Help me out please!

Did you just jump straight up to 1.45 from 1.35? If so back off down to 1.35 again and start increasing in the smallest increments you can on the vcore, until you're prime stable. My chip won't even boot to windows at 4.8 and 1.35 but is stable at 1.4, so you'll probably end up running a lower vcore than me; well clear of 1.52!

I tried a myriad of voltage settings including, but not limited to:1.2 (BSOD)1.3 (BSOD)1.35 (Boot @ 4.8/ P95 instant crash)1.375 (Boot @ 4.8/ P95 instant crash)1.390 (Boot @ 4.8/ P95 instant crash)1.4 (Boot @ 4.8/ P95 not so instant crash)1.42 (Current Setting) Stable in P951.45 (Stable in P95)1.51 (Offset going crazy, turned off PC when reading came.)Settings- Current Working config -BCLK - 100Multi - x48PLL OverV - OnEnergy Save - OffDRAM timings set to specsLoad Line Calibration - HighVRM Manual - 350Phase & Duty control - ExtremeCPU Voltage - Manual - 1.42DRAM v - set to specVCCSA - 0.9VCCIO - 1.0

What you've got there already looks ideal. You could try bumping the PLL to 1.8125 and the VCCIO to 1.2, then slowly back off the vcore to see if it remains stable. Also set your RAM to 1333/auto during this to remove it from the equation. Try with LLC set to extreme as well as I found that my vdroop scaled badly to mult increases. Could be that even though you've set 1.4 in bios, the chip is only getting 1.38 under load.It may just be that you're in a similar boat to myself, and have gotten a mediocre chip. If you are stable at 1.42 then tbh I would probably just leave it there if temps are fine. Nobody really knows what sort of voltage these chips can take yet so everyone who's clocking is taking a risk no matter what voltage they're putting in. Really sucks when I hear of the people getting 5Ghz+ on 1.35v, but there's not much we can do about it! Well short of degrading the chip deliberately and getting it replaced on warranty :Whistle:

Are those voltages the ones you set in BIOS or full load Vcore? If it's 1.42 BIOS Vcore + High LLC, you probably have a little less Vcore than that at full loadAlso, did you enable CPU/PLL overvoltage? it might help at 4.8GHz

Set in BIOS, PLL Ov is on, otherwise it will freeze while loading windows.P95, with 1.42 gets the max temp readings of 89, with an average of 70-88 depending on the test.

What you've got there already looks ideal. You could try bumping the PLL to 1.8125 and the VCCIO to 1.2, then slowly back off the vcore to see if it remains stable. Also set your RAM to 1333/auto during this to remove it from the equation. Try with LLC set to extreme as well as I found that my vdroop scaled badly to mult increases. Could be that even though you've set 1.4 in bios, the chip is only getting 1.38 under load.It may just be that you're in a similar boat to myself, and have gotten a mediocre chip. If you are stable at 1.42 then tbh I would probably just leave it there if temps are fine. Nobody really knows what sort of voltage these chips can take yet so everyone who's clocking is taking a risk no matter what voltage they're putting in. Really sucks when I hear of the people getting 5Ghz+ on 1.35v, but there's not much we can do about it! Well short of degrading the chip deliberately and getting it replaced on warranty :Whistle:

Just upped the PLL to 1.8125, didn't touch the VCCIO as it went pink... so I didn't know is it was safe of not.. you tell me. Set the LLC to extreme... toned down Vcc to 1.375.P95 is behaving... not... failed after 3 minutes of testing.

Sorry that I missed the PLL overvoltage line in your previous post.What full load Vcore do you have? the BIOS setting is irrelevant.And your temps are too high Ioan. Maximum 80ºC and 1.38V (full load) in my book

Sorry that I missed the PLL overvoltage line in your previous post.What full load Vcore do you have? the BIOS setting is irrelevant.And your temps are too high Ioan. Maximum 80ºC and 1.38V (full load) in my book

Currently, on P95.. 1.408Temps hovering in the 70 with occasional jumps to the 80. Max recorded 89. I just don't get it.. how can I run stable at 1.38?

Currently, on P95.. 1.408Temps hovering in the 70 with occasional jumps to the 80. Max recorded 89. I just don't get it.. how can I run stable at 1.38?

Try turning HT off otherwise I think you will struggle to run at 1.38V with HT on unless you have a really good chip.

that value's proximity to the 1.52 "Official" limit is worrying me. Help me out please!

Just a heads up that 1.52v is reference to "vid" not vcore. There is no direct correlation to vcore represented by vid. Read foot note 2 of the specification sheet: "Each processor is programmed with a maximum valid voltage identification value (VID) that is set atmanufacturing and cannot be altered. Individual maximum VID values are calibrated during manufacturingsuch that two processors at the same frequency may have different settings within the VID range. Notethat this differs from the VID employed by the processor during a power management event (AdaptiveThermal Monitor, Enhanced Intel SpeedStep Technology, or Low Power States)."I see so much misrepresentation of that spec out there that it isn’t funny.

Just a heads up that 1.52v is reference to "vid" not vcore. There is no direct correlation to vcore represented by vid. Read foot note 2 of the specification sheet:"Each processor is programmed with a maximum valid voltage identification value (VID) that is set atmanufacturing and cannot be altered. Individual maximum VID values are calibrated during manufacturingsuch that two processors at the same frequency may have different settings within the VID range. Notethat this differs from the VID employed by the processor during a power management event (AdaptiveThermal Monitor, Enhanced Intel SpeedStep Technology, or Low Power States)."I see so much misrepresentation of that spec out there that it isn’t funny.

Oh so that's what VID meant... (Please forgive me i'm new at this.)For know i'm staying at 1.42v Highest temp in P95 is 89, but it doesn't stay long there. Stays mostly in between 70-80. Ran Crysis out of curiosity yesterday and I am mind blown. Its playing like water all maxed out... with 16xQ :o (I know its the 580's play here but still, DAMN!) And the max cpu temp was about 64.Running the final P95 stress test for 3 hours, then will install W7 clean (way to many BSOD's and incorect shutdowns in my OC path.3 hours is enough right?And 1.425 is fine right? CPU wont die of overvoltage right? :|

It fails after 20 minutes with a stopped worker. It this considered stable?@ 1.425v

If it fails at all, it's not stable. You've tried about everything so you're just going to have to accept that you have an average/bad chip for clocking and aren't going to be able to run 4.8GHz at under 1.4v on the core. Either knock the mult down to 47 and run things at that speed on a lower vcore, or make a choice to start venturing into much higher voltages for that extra 100MHz. Not worth it imo but it's your call. I'd also be VERY careful about reaching temps as high as 90 as that's really getting to the upper limits of things. Would be well worth investing in a much better cooler if you want to keep pushing.

If it fails at all, it's not stable. You've tried about everything so you're just going to have to accept that you have an average/bad chip for clocking and aren't going to be able to run 4.8GHz at under 1.4v on the core. Either knock the mult down to 47 and run things at that speed on a lower vcore, or make a choice to start venturing into much higher voltages for that extra 100MHz. Not worth it imo but it's your call.I'd also be VERY careful about reaching temps as high as 90 as that's really getting to the upper limits of things. Would be well worth investing in a much better cooler if you want to keep pushing things.

Might be the way to go, 4.7 GHz is the last mult value before I need PLL Ov to even load windows. Do you know any optimal voltage for 4.7 below 1.4v?Currently trying a last ditch effort to run on 1.375 but with the LLC to Extreme it's at 1.4v .. now 1.392.. Test 4 waiting the failure... there it is worker 4 fails..Sigh...