OCUK Sandy Bridge, SB-E and Ivy Bridge 5GHZ Club

SimonMaltby · 13 Jan 2011 at 18:34

EniGmA1981 said:
5.3??? Wowser lol

Quick question, is high voltage dangerous refardless of temperature increase? that is, will just the voltage frazzle the chip regardless of cooling? Even if it was theoretically running loaded at 20*c?

Just to put a bit of perspective here...

The guidance about voltages is just that guidance and it was also mentioned when Sandy Bridge had just launched and One chip (mine a 2500K) stopped working properly. We still don't know why it stopped working and probably never will.

As the week has gone on OCUK have stress tested some Sandy Bridge processors at higher voltages and nothing has failed yet.

Of course high voltages are risky, that's why you take it slow and easy, step up very gently and stop when you feel uncomfortable. Running these voltages day in day out would be very likely to cause problems, but for some benchmarking and screen shots to brag about, I think they are fine.

Having said all that i'm quite reluctant to go above 1.45 vcore.

DBT85 · 13 Jan 2011 at 20:29

SimonMaltby said:
Having said all that i'm quite reluctant to go above 1.45 vcore.

I'd have to agree. For the bump up i needed just to get stable long enough to run super pi it just isn't worth it for me. Possibly in the fullness of time someone will find out that setting xyz to 1.523542352v will give you some more headroom.

For the moment though its still all on the vCore and at that, I'm happy at 1.38. At the moment that looks like it might mean a max clock of 4.6 for me at the moment though at 4.7 BSOD 2 hours into Prime.

bulldog147 · 13 Jan 2011 at 21:32

SimonMaltby said:
Think I might be getting a little carried away now....

How gutting to be so close to a sub 7s time

Your cpu will reduce in lifespan sooner. Cos one day OCUK and myself will going to say "told ya so!".

cmsbfent · 13 Jan 2011 at 21:46

How are the other benchmarks coming along?

People deserve to know!!

SimonMaltby · 13 Jan 2011 at 21:46

Ok final SuperPi test I'm going to run... enough is enough

Now to get that magic 5GHz nice and stable and do some graphics benchmarks

Subliminal Aura · 13 Jan 2011 at 22:38

matthab said:
This could be ive bricked my Sandy Bridge Club .

Yep ... subscribed.

The 1.45v Fry Up Club

el Kuarlos · 13 Jan 2011 at 22:47

I think intel might have shot themselves in the foot making these chips run so cool- least with hotter ones you know when to stop...with these it seems to bork before the temps get unsustainable!

DBT85 · 13 Jan 2011 at 22:52

SimonMaltby said:
Ok final SuperPi test I'm going to run... enough is enough

Now to get that magic 5GHz nice and stable and do some graphics benchmarks

Whats going on there? Asus prog says 1.365v and CPUZ says 1.464.?

Speed · 13 Jan 2011 at 22:52

el Kuarlos said:
I think intel might have shot themselves in the foot making these chips run so cool- least with hotter ones you know when to stop...with these it seems to bork before the temps get unsustainable!

Not really, there isn't really a downside to lower heat output on any computer component. The main problem is simply the fact users are so used to clocking 45nm CPUs that many don't seem to realise that using the same sort of voltages on 32nm process CPU is a receipe for disaster.

Chuggerboom · 13 Jan 2011 at 22:57

Fry your chip thread more like. Even Overclockers UK state not to go over 1.38v.

Anything over 1.38v will limit CPU lifespan and anything over 1.42v will likely kill the CPU or severely limit its lifespan.

http://forums.overclockers.co.uk/showthread.php?t=18227651

bulldog147 · 13 Jan 2011 at 23:01

SimonMaltby cpu could dies anytime sooner or later if he kept pushing more than 1.45vcore as he probably think ocuk will replaced his dead cpu if anything has happen to him.

SimonMaltby · 13 Jan 2011 at 23:03

Todge said:
Whats going on there? Asus prog says 1.365v and CPUZ says 1.464.?

I booted at AUTO vcore in bios and setting the vcore in windows if bios in Auto makes them not match... the CPUz value is the correct one.

SimonMaltby · 13 Jan 2011 at 23:09

bulldog147 said:
SimonMaltby cpu could dies anytime sooner or later if he kept pushing more than 1.45vcore as he probably think ocuk will replaced his dead cpu if anything has happen to him.

Overclocking is a risk. My first CPU failed, but not because I did any stupid overclocking. This one is being overclocked in a controlled and calculated manor. I would not dream of running these voltages for long and would not dream of returning it for an RMA if if failed and i thought it was because I pushed it too far.

I brought this chip to have fun, and to me fun is benchmarking. I know and accept the risks as do many others on the forum.

There are many levels of overclocking. Here they mainly fall into two types: like me that want to get the best benchmarks they can, those that want a safe extra bit of speed.

So chill a little and just enjoy seeing some nutter go mad LOL.

Chuggerboom · 13 Jan 2011 at 23:14

SimonMaltby said:
Overclocking is a risk. My first CPU failed, but not because I did any stupid overclocking. This one is being overclocked in a controlled and calculated manor. I would not dream of running these voltages for long and would not dream of returning it for an RMA if if failed and i thought it was because I pushed it too far.

I brought this chip to have fun, and to me fun is benchmarking. I know and accept the risks as do many others on the forum.

There are many levels of overclocking. Here they mainly fall into two types: like me that want to get the best benchmarks they can, those that want a safe extra bit of speed.

So chill a little and just enjoy seeing some nutter go mad LOL.

That is up to you, after all it's your risk. Both Overclockers UK and Intel say that going over 1.38v is not recommended. Happy frying

bulldog147 · 13 Jan 2011 at 23:15

Fair point, But again, it at your own risk! When I brought intel i7 980x from ocuk I didn't go over 1.38v because i cannot risk it.

Enjoy your biggest overclocking! See if you can go all the way 6GHz !

reygjabla · 14 Jan 2011 at 00:03

SimonMaltby said:
Ok final SuperPi test I'm going to run... enough is enough

Now to get that magic 5GHz nice and stable and do some graphics benchmarks

Looks like 5.5 GHz is attainable on this particular chip. :cool:

SimonMaltby · 14 Jan 2011 at 00:09

Ok, at last the 5GHz Prime 95 1 Hour stable is achieved.
Took a little more vCore and PLL

That's me happy.

Not about tomorrow, so Graphics benchmarks etc at 5GHz will have to wait till the weekend. I think I have done quite enough for today

cmsbfent · 14 Jan 2011 at 00:17

SimonMaltby said:
Not about tomorrow, so Graphics benchmarks etc at 5GHz will have to wait till the weekend. I think I have done quite enough for today

Well thats not good enough! Get some coffee in you and man up!

Gerard · 14 Jan 2011 at 00:23

Westyfield2 said:
Nope. LGA2011 for that I believe.

Though I'd bet you'd hardly see the difference between 16x 16x and 8x 8x anyways .

Pretty sure the maximus has 2 slots that run at 16x and one at 8x. Makes bugger all difference ingame anyway.

Westyfield2 · 14 Jan 2011 at 00:25

Gerard said:
Pretty sure the maximus has 2 slots that run at 16x and one at 8x. Makes bugger all difference ingame anyway.

Yeah as Todge said, Nvidia nf200 chip onboard to add more lanes.