i7 5930K, safe voltage with safe temp.

Probably the Asus board applying more voltage than the EVGA one.

What is set in the BIOS is one thing, what the CPU actually receives is another (and may be board dependent), is my guess.
 
Necro thread GO!

Here's an odd one. For sundry reasons I had to swap motherboards was running on an X99 Classified. Easy peasy overclock (results above). Enable XMP, set multiplier. Done.

Anyway, moved bits over to Asus Strix X99. All hardware identical, down to the thermal paste (cleaned up and re-applied, of course). Runs 4C hotter idle. ¯\_(ツ)_/¯

Of course had to set it all by hand since Asus' "automatic" stuff is about as useful as a dead raccoon.

It depends on what you mean "all by hand" since even a small setting may create the temp difference
 
Newbie here.

Old thread I know but I though my i7-5930k on an ASUS X99 Pro was a dud and would only overclock to 4.2@ 1.30v at the bios. BCLK @ 100.00 MHz. No luck, it was stuck at 4.2MHz

I did a little tinkering in AI suite 3 (I know, it's not supposed to do much for overclocking) by raising the voltage to 3.3v then lowering it back down to 1.312v. It worked! I've now got a stable 4.5MHz

Temps are maxing @ 61C during a prolonged AIDA64 stress test using an Alphacool Eisbaer 240 Liquid cooler.

Happy puppy indeed :)
 
Last edited:
Old post, but...

I've been OC'ing my 5930k for 5+yrs now and am beginning to wonder if I f'd up.
It's at 1.452V and 4.3 ghz w/ a 166fsb w/ Ram at 1317 Mhz. Runs HOT. Doing FEA in Solidworks, which when on the gas has the package peaking at 90°C.
Water cooled with a H100i


I'm wondering if I should back off the voltage some and seek out some more multiplier...???

It's my CAD station which is getting an upgrade to a 9900k setup in the next week or so, but this one machine will trickle down to my daily driver.
 
I've been OC'ing my 5930k for 5+yrs now and am beginning to wonder if I f'd up.
It's at 1.452V and 4.3 ghz w/ a 166fsb w/ Ram at 1317 Mhz. Runs HOT. Doing FEA in Solidworks, which when on the gas has the package peaking at 90°C.
Water cooled with a H100i

That is a LOT of voltage. The general rule that I've always heard, with regards to Haswell-E, is to have good cooling if you go over 1.3v, and especially over 1.35v. Never go over 1.4v no matter what. 1.45+ really should not be required, especially at only 4.3Ghz, and is high enough that it might even cause CPU degradation over time. Perhaps you just got a really bad chip. You might try adjusting the LLC / VDroop settings in your BIOS instead of cranking up the voltage so high. I run my 5820K at 4.5Ghz using 1.375v. I'm not surprised that you are seeing high temperatures, even with water cooling.

It seems very odd to be running 166Mhz FSB, especially with an unlocked processor where all you have to do is raise the multiplier in order to overclock. The only reason people generally have to change the FSB is to achieve a specific RAM frequency, but between the 100Mhz and 125Mhz FSB settings, you should pretty much have all RAM speeds covered. Why are you running 166?

Finally, it might be worth mentioning that when I built my 5820K setup, I used an H100i GTX with 4 fans in push-pull and used that for years. Eventually my temperatures started to skyrocket, to the point of having BSODs, and I tried everything I could think of to figure out why. Turns out that over half of the fluid in my radiator had evaporated, which obviously destroyed it's ability to cool the CPU. I replaced it with a different AIO and everything was good again. I'm not sure what version of the H100i you have, but considering it's age at this point, it might be worth it to check on it's health and make sure you aren't in the same boat that I was.
 
Yea, I read this thread and started wondering. I'm back in performance mode and OC'ing. Like I said, this OC was from 5yrs ago. It runs Solidworks/FEA stabile. Gets into a cycle w/ all cores nearly maxed for a good while. Hot little bugger tho.
I'm thinking I need to take another look at what I did to OC this chip...after I finish this FEA job that has 24hr runs involved.

Same H100i GTX you had/have. How do you check the coolant level? I see no reason to dispose of it if it's just a little low. Seems to be working. I have a new H100i Pro that I'm going to put on the 9900k build.

I'm running the high FSB to get the memory pathway as fast as possible.
 
Still have my i7 5930k running at 4.5GHz, 1.335v on the core. Been this way since maybe April of 2016? Full custom watercooling setup though with 4x120mm external radiator (EKWB). Runs idle in the 30's, and max load in the low 70's during OCCT/Prime torture tests for 24hrs.

ZlSwls4.png

MipflRS.jpg
 
I built my work pc in 2014 when these chips came out with an H100i and it still works fine there on my 5820k. I did bump up the OC from whatever it was before (either 40x or 42x) to 44x at 1.28v a year or two ago but I haven't had any issues. I have old 3000 kingston hyperx memory but 125 divider sucks on my CPU so I run it at 2666/100.
 
Same H100i GTX you had/have. How do you check the coolant level?

Well the easiest way is to simply unscrew your radiator from your case, and while it's above the hoses and pump, move it around and check if you can feel water sloshing around inside the radiator. That usually means there is air inside. In my case, so much coolant had evaporated that it was also noticeably lighter than it should have been.
 
All I know is that the machine is running FEA hard. Like full throttle for 30+hrs. The package is reaching 90C. And it's fine. Been doing this for years.
I'll definately be looking into the cooler though before it's repurposed once the 9900k is built
 
Last edited:
The package is reaching 90C. And it's fine. Been doing this for years.

90C package temp is REALLY hot. Your CPU might actually be throttling at that point. I know that with my 5820K @ 4.5Ghz, if even one single core hits 85C, I will get a WHEA_UNCORRECTABLE_ERROR BSOD like clockwork. I know my OC is stable because as long as I stay below that temp everything is great. So that's pretty amazing that you can even get yours to 90C.
 
I get no BSOD. It just runs hot. I know I can crank it up another notch boot and run w/o error, but then it runs tasks slower (measuring with a stopwatch.)
 
It's perfectly safe to run a Haswell-E CPU @ up to around 1.35v. I have a 5960X that was clocked to 4.5GHz that I beat the shit out of for a couple of years like that. I have another that did 4.4GHz for over 5 years at 1.3v.
 
Back
Top