sandybridge take note

beware those with sandybridge linked is a thread from overclockers uk
http://forums.overclockers.co.uk/sho...php?t=18227652

Hi there


Right guys myself and our technical guys have spent the entire weekend and this morning in discussions with Intel regarding the alarming amount of reports of Sandybridge CPU's dying and have been conducting our own testing as have Intel to find out what is a definite no no.


Sandybridge maximum safe voltages

Core Voltage - Not recommended too exceed 1.38v, doing so could kill the CPU, we therefor recommend a range of 1.325-1.350v if overclocking.
Memory Voltage - Intel recommend 1.50v plus/minus 5% which means upto 1.58v is the safe recommended limit. In our testing we have found 1.65v has caused no issues.
BCLK Base Clock - This is strictly a NO, anyone using base clock overclocking could/will cause damange to CPU/Mainboard. (Set manually to 100)
PLL Voltage - Do not exceed 1.9v!!



Processor - Basically we recommend customers not to exceed 1.35v to play it safe, all our bundles are set at 1.3250v or lower, any competitors offering bundles above 4.6GHz you should be enquiring as to what voltage they are using as we believe anything over 1.38v will limit CPU lifespan and anything over 1.42v will likely kill the CPU or severely limit its lifespan.

Memory - Intel recommend 1.50v plus/minus 5% which means 1.60v is the ideal safe maximum, but we have found in our testing all 1.65v memory is fine. We have also found most new 1.65v like Corsair XMS3 will run at its rated timings with just 1.50-1.55v which is well within Intel specifications. So people upgrading to Sandybridge you can still use your old DDR3, but we do recommend you run it at 1.60v or less. We are shipping most of our bundles which feature Corsair XMS at 1.50v-1.55v at rated timings. We've also discussed with Asus and MSI regarding voltages for memory and they also confirm in their testing 1.65v caused no issues with reliability.

Base Clock - To put it simple if you value the life of your components, do not overclock using base clock!

PLL Voltage - Again do not exceed 1.9v!

Yes, I won't be going over 1.38vcore once I get my 2500K. Don't want to kill it, lol.

Interesting.. Good thing to see that RAM can (keeps fingers crossed) can be run at 1.65v. Would the ram being able to run below spec @1.5V be based solely on the memory's ability to do so or could the SB IMC have an impact on that?

LOL ok so you talked about it for a couple of days and you came to XS to tell these guys what not to do without any solid proof. You do realise you will just entice more people to prove you wrong now :D. Intel told us the same thing for all the CPUs in the past 3 years but there is not much to support that for example memory above 1.5 kills CPUs, or VTT above 1.4 or whatever.


thanks for the feedback though, do you have some numbers for us of people that have killed and actual situations that cause chips to die, for example ive seen someone on that forum specifically recommend NOT to oc with bclock at all or chips will die but based on what evidence or information? I dont know if D1 are different but mine havent died after a couple of months of investigative OCing hehehe

I agree. I'm cranking the Base clock to 107 and running 1.53 on V Core.

While I think that 1.35vcore is a smart move for long term 100% usage it's hard to agree with you based on what we've seen here.
Benches at up to 1.6vcore( I shuddered at that) and on air.
Personally for 100% 24/7 usage I'm going to take what I can get at 1.375vcore and for the few benches I do stay under 1.5vcore

Ok,so let me tell you my story.Had a i3 560,sucker was very poor,4 ghz 1.37 stable.Beched some dice at it,managed to get only 5.57 for pifast and some 5.4 for superpi.So i decided that i will have some fun and try to kill it to get it to rma and get other piece that clocks better.Here is the fun part.I had 2 sessions with it on DICE,it scaled to 1.9V.So it run about 15 hours on 1.9v,get it on air,ran some prime at 1.7v(it was maximum i could get in load,crappy board ud4) about an hour,then tried 1.6 and 1.7 vtt.Still nothing.Again anoher dice session with 1.9V,huge qpi voltage.Result?Nothing,the sucker not even degraded a bit,absolutely nothing.Few days ago i benched with 2500K for 8 hours at 1.65V core and 1.85mem(on a crappy air cooler).Everything went well.

Do i believe in your story OP?

No i don't.About the voltage,quick question.Which one is safer 1.4V with 75C load or 1.5 with 45C load? I bet my money on the last one.;)

I dont think anyone is going to really see anything special by going from say 4.5 to 5Ghz anyways and most CPUs will probably do 4.5 at stock voltage which is what you want for a 24/7 setup anyway but i bet you anything people here and everywhere else will be obsessed with 5GHz so much like everyone was with 4Ghz for a long time with Q6600s and later new chips 920s 750s etc

Nice interest

OK Alex-Ro you're officially crazy just like my good mate Uncle Fester :rofl:

Voices inside my head tell me i'm healthy :D Few people also told me when i benched with dice on 31 december at 10pm :D

Quote:

---

Originally Posted by dinos22

Intel told us the same thing for all the CPUs in the past 3 years but there is not much to support that for example memory above 1.5 kills CPUs, or VTT above 1.4 or whatever.

---

Exactly. Ran my i7 on 1.47vtt for a solid year on water until it died haha.

Time to run at least 1.45 vcore on my incoming SB for 24/7.

Looks like I'll be slaughtering chips left and right under water 24/7 at these numbers.

Wish the "shorten the lifespan of your CPU" crap would go away and stop scaring noobs away from OCing :p:

Quote:

---

Originally Posted by dinos22

I dont think anyone is going to really see anything special by going from say 4.5 to 5Ghz anyways and most CPUs will probably do 4.5 at stock voltage which is what you want for a 24/7 setup anyway but i bet you anything people here and everywhere else will be obsessed with 5GHz so much like everyone was with 4Ghz for a long time with Q6600s and later new chips 920s 750s etc

---

Boy did you take the words right out of my mouth.
5GHz will be everyones goal but I can understand that.
They see benches from "talented"OC'ers showing 5.4.-5.5 and everyone builds the system and before ever booting to windows sets the bios at 5GHz..:rofl:
If I can offer anything it's that these chips will have variables.
We'll see the dogs that need 1.4v for 4200 and the gems that will do 5.5 at under 1.4v..
Learn the chip,see what it does and slowly bring it up.
Be smart here and be happy with what your chip does and forget what anyone else gets except for general knowledge for your own testing.

5Ghz may look normal for 2500K/2600K

Come on 1.4 for 4200 it is not possible for some chips to be so crappy.Maybe some chips will have trouble reaching 4500,but 4200 should be easy with all.

Sorry I just plain loll'ed, this should be a sticky in ModerateSystems forum

Ofcourse Intel guys will tell you not to exceed certain values... sigh...

Just cool it adequately and just be wise ( as stated by the two gods aka spammers of this forum ) when bumping volts. Just find the sweetspot for your setup and live with it... or swap, sell and buy till you find a gem... pretty happy with my current setup at 4.5 at 1.25volts...

Quote:

---

Originally Posted by dinos22

I dont think anyone is going to really see anything special by going from say 4.5 to 5Ghz anyways and most CPUs will probably do 4.5 at stock voltage which is what you want for a 24/7 setup anyway but i bet you anything people here and everywhere else will be obsessed with 5GHz so much like everyone was with 4Ghz for a long time with Q6600s and later new chips 920s 750s etc

---

One need a few good obsessions.
:yepp:

Quote:

---

Originally Posted by Movieman

Learn the chip,see what it does and slowly bring it up.
Be smart here and be happy with what your chip does and forget what anyone else gets except for general knowledge for your own testing.

---

:p: amen to that brother, that is my approach also.
even if people do better or worse, i still try and manage my own.
i am not very experience when it comes to OC and all, but i sure
do learn a lot from the content provided by other users.

like i know i see benches of people with my same cpu running ~5Ghz and all.
at the same time i ask myself do i need to run that?
do i need to do an ice/ln2 session to achieve 5Ghz even if just for a day?
or do i settle for my 24/7 4.2Ghz/2000mhz on the ram. which is only 7-10% less than 5Ghz.

i guess what im trying to say is, we must learn how to live with ourselves.

What kind of muppet would come here and post that garbage? Keep it to the ocuk fanboy forums plz! :lol:

Interesting to note the 0.5v vccio / vddr limitation is still there. If cpu's are dying that is likely the cause. People don't seem to think they need any kind of knowledge about what theyre doing besides raising voltages and multipliers. Blundering straight in to a maximum overclock on a new cpu is just asking for problems (as MovieMan pointed out above). The more experienced people here are having no issues with running silly vcore as we've seen on the threads here over the last few weeks. I see no dead cpu's here?

overclockers.co.uk are full of :banana::banana::banana::banana: anyway :p

Harrumph.

Quote:

---

Originally Posted by Movieman

They see benches from "talented"OC'ers showing 5.4.-5.5 and everyone builds the system and before ever booting to windows sets the bios at 5GHz..:rofl:

---

+1
I am still quite the noob because that was my plan :rofl:. At least by the 2nd or 3rd boot.

So, is is it really safe to run RAM @ 1.65v?

Quote:

---

Originally Posted by OC Nub

+1
I am still quite the noob because that was my plan :rofl:. At least by the 2nd or 3rd boot.

---

:rofl:That was a good one. I agree with serious testing and that is just one reason why we are here anyway,Xtreme means just that. I build all my systems with xtreme cooling and exercise caution when testing stability but always shoot for the max. Go big or go home.:up:

Quote:

---

Originally Posted by Movieman

Boy did you take the words right out of my mouth.
5GHz will be everyones goal but I can understand that.
They see benches from "talented"OC'ers showing 5.4.-5.5 and everyone builds the system and before ever booting to windows sets the bios at 5GHz..:rofl:
If I can offer anything it's that these chips will have variables.
We'll see the dogs that need 1.4v for 4200 and the gems that will do 5.5 at under 1.4v..
Learn the chip,see what it does and slowly bring it up.
Be smart here and be happy with what your chip does and forget what anyone else gets except for general knowledge for your own testing.

---

Wise words from Mr. Movieman :up:

I feel as though yes there can be processor degradation but from what i heard most of the dead CPUs were from too much BLCK, which I still don't quite understand. Now in the past Intel recommended 1.375v for bloomfield and then vabsolute was 1.55v. Now I don't understand why intel has to publish a datasheet and just say VID upto 1.52v. From what i take with that is if you keep voltage at auto vid will change upto 1.52v when processor is OCed to a very high frequency like 5-5.2ghz. i have actually seen my VID goto 1.475v at 5.2ghz. i did see voltage degradation from 1.53v and then the week after i needed 1.56v and then 1.58v, now that was at 5.2ghz on a D1 stepping chip. i did make note of it in my article, those are with LLC added. From what i saw was that the temperatures on air were very high at auto, and it is well known that heat and voltage will kill a chip. From talking to a lot of people I have gathered that they do not just die because of nothing, extreme overvolting and heat can result of the death of any chip. Companies like MSi that have published OCing guides have stated that high voltage is 1.55v. Now i understand that 1.5v is a lot to fathom for a 32nm chip, but heck so is 5+ghz.

it does seem like retail chips are much weaker than initially thought, their average OCes are not 5ghz, more like 4.5-4.8.

Quote:

---

Originally Posted by Movieman

If I can offer anything it's that these chips will have variables.
We'll see the dogs that need 1.4v for 4200 and the gems that will do 5.5 at under 1.4v..
Learn the chip,see what it does and slowly bring it up.
Be smart here and be happy with what your chip does and forget what anyone else gets except for general knowledge for your own testing.

---

I like all my chips to be nice clockers... a plain "ole" average chip just doesn't cut it... to be sold off at the feed lot.

Quote:

---

Originally Posted by sin0822

it does seem like retail chips are much weaker than initially thought, their average OCes are not 5ghz, more like 4.5-4.8.

---

I'm finding that with my chip as well. 4.7 is pretty close to stable, 4.8 isn't stable yet.

Makes me happy to see you say that too, I've been over your OC guide so many times now asking myself, what the hell am I doing wrong ?

Quote:

---

Originally Posted by PolRoger

I like all my chips to be nice clockers... a plain "ole" average chip just doesn't cut it... to be sold off at the feed lot.

---

That's ok but you can find out if they aqre gems or dogs without killing them.
That's really the point I was trying to make.
For the other guy just topped 5GHz in 3 mins after initial boot up.
Ran SP1m at 5088mhz.
Seems to be stable at 4588mhz on 1.35vcore.

yes just like with my 5.2ghz chip i can run spi, but i have to drop down to 5.1 for benches with .6v less, b/c of heat. The 4.8ghz chip will do 4.8ghz and bench 4.8ghz it has room with voltage, but I think no CPU PLL for D1 is killer. Maybe Intel is afraid that it would make their retail D2s look bad, which they are doing on their own. I have been looking around very busy places looking for SB OCes, maybe 1 over 5ghz, one at 5ghz, and a lot under 5ghz, more between 4.5-5. Here people are hitting over 5ghz no problem it seems tho, but this is one place that real OCers gather.