• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

max vccin?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Tír na nÓg

Member
Joined
Sep 19, 2015
Hi lads,

I was wondering how high I cold take the VCCIN on an Haswell-E.

It used to be 2.2v opn Haswell, but I read here and there that 2v is the mac for Haswell-E (which seems odd...).

The rule is to have +500mV on the VCCIN than on the vCore, but it doesn't exactly work lije that, right?

As an example: I run the 5820k in [email protected]/1.32v/2.15v VCCIN. If I go any lower than 2.15v, I lose stability.

As I've been lucky enough to get a good chip, avoiding degradation in the next couple of years would be not so bad!

What's your input on this?
 
One thing I do know is that it takes more for the 5960x than it did for my 4790k. I don't run mine 24/7 but usually it's stable around 1.9 at 4.5G. Once I put the gas to it though it jumps pretty quickly. IMO if you want to keep the chip around for 10 years then run it on auto at stock. If you're willing to push it a bit then you know the consequences. Possible premature degradation but kept within reason shouldn't affect longevity that much. 2.1 is most likely fine as long as the temps are down. If you're running at 80+ all the time then..............
 
One thing I do know is that it takes more for the 5960x than it did for my 4790k. I don't run mine 24/7 but usually it's stable around 1.9 at 4.5G. Once I put the gas to it though it jumps pretty quickly. IMO if you want to keep the chip around for 10 years then run it on auto at stock. If you're willing to push it a bit then you know the consequences. Possible premature degradation but kept within reason shouldn't affect longevity that much. 2.1 is most likely fine as long as the temps are down. If you're running at 80+ all the time then..............

Man, in 10 years, I'll be dead!

It maxes out@71c-81c through all cores with P95, and usually runs in the high 50's when using anything more demanding than Firefox and VLC.
 
Man, in 10 years, I'll be dead!

There's our answer ! Seriously though, there are probably very few people who know the absolute max for VCCIN as you said .5v higher than core. Most of them are using LN2 and that's a whole new game.
 
[email protected]/1.34v - Cache 4430/1.33v - VCCIN 2.18v

I wait until tomorrow morning to see if that's stable (3 hr in p95 right now) and I'll call it a day...

And might finally use the rig for something else than stability testing and benchmarking :p

Thanks Johan
 
What happens if you drop the VCCIN and up the V_Core?
 
Giving a try right now: dropped the VCCIN to 2.05 and upped the vCore to 1.35v.

Will let you know tomorrow morning how it went.

Edit: I have to let p95 run for a good 2 hours at least, as I got some 101/124 BSOD's after 1 hour or more.
 
You set higher vccin only when CPU is losing stability even when you raise vcore. I was benching at 5.5GHz+ with vccin 2.2V. You shouldn't need more than 2V for sub 5GHz clock. Vccin above 2.2V may kill or degrade your CPU ( usually above 2.3V but every CPU is different ).
Remember that internal CPU temp is different than what sensors are registering. Also higher CPU clock = lower tolerance to temperature. So 80*C can be fine at 4.5GHz but not at 4.7GHz. Actually I had like that on my 5820K. By lowering temp about ~30*C I could make next 200-300MHz at the same voltage.
 
^ Absolutely agreed on the temp impact:
<80c=>4.85GHz/1.335v
>80c=>4.85GHz/1.35v

Finally got it [email protected]/1.315v/VCCIN 2.15v

Anything below 2.15v and it crashes. Passed p95 overnight with the above. Temps in the high 70's.
 
Well, little update: it seems I have been a bit optimistic on the cache...

I lowered it to 4.2GHz/1.28v. Decreased the VCCIN to 1.95v. [email protected]/1.335v, and everything seems OK (running P95 for 25 min now).
 
Back