• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

The Penryn Voltage Thread

P

Pilot53

Limp Gawd
Joined
Nov 19, 2006
Messages
249
There is alot of nervous talk going around about the new 45nm penryn cpus not being able to take alot of voltage. Right now people are saying to leave the vcore at 1.4 or less, 1.45 being the max. Some people have even reported failures at around 1.35v. It seems like heat is not going to be the limiting factor when overclocking penryn, but the voltage, pretty much no need to watercool a penryn since heat isnt goign to limit the oc. I think we should have a thread where people can post their experiences with penryns and any info they have about penryn voltages.

Here is the official info from intel:

At conditions outside functional operation condition limits, but within absolute
maximum and minimum ratings, neither functionality nor long-term reliability can be
expected. If a device is returned to conditions within functional operation limits after
having been subjected to conditions outside these limits, but within the absolute
maximum and minimum ratings, the device may be functional, but with its lifetime
degraded depending on exposure to conditions exceeding the functional operation
condition limits.

At conditions exceeding absolute maximum and minimum ratings, neither functionality
nor long-term reliability can be expected. Moreover, if a device is subjected to these
conditions for any length of time then, when returned to conditions within the
functional operating condition limits, it will either not function, or its reliability will be
severely degraded.

1.45v is the absolute max, similar to the absolute max for 65nm of 1.55

Obviously intel has to protect its interests, so there is a slight chance that these values are fairly conservative and we might be able to get away with more and still have our cpu last until the nest product cycle, but that it the point of this thread to uncover information about how much voltage these cpus can take, and to warn people with 45nm cpus not to apply the same voltages they have been comfortable using with 65nm.
 
Every new process has its voltage limits..

90nm --> 1.8v
65nm --> 1.5v
45nm --> 1.4v

Seems like a logical progression to me...
 
commodore said:
Every new process has its voltage limits..

90nm --> 1.8v
65nm --> 1.5v
45nm --> 1.4v

Seems like a logical progression to me...
Click to expand...
It really does make sense but there has also been talk of a "burn-in" which I'm kind of skeptical of but people are throwing that idea around on XS. So, time will tell. One thing is for sure, these things have different limits that'll have to be discovered.
 
Yeah XS has some pretty worrying posts about this, but some people might have found a possible link with SOME premature failures and dfi boards. This situation is also rendering watercooling useless with dual cores, unless your after the "bling" or silence, im sure I could get the same overclock on air that i have with water, keeping the voltage that low on a 45nm pretty much eliminates the need to combat heat for most of us to hold a stable safe overclock. Time will tell, it would sure be nice if we find out that these chips can take a little more, since they are responding well to voltage and running cool in the short term.
 
no one else has anything to say about this problem? It seems a lot of sites I go to talk about '4.0ghz+ with over 1.4v'

If they are volting that high, and some people are saying there are electromigration problems, I am confused why this is not the #1 topic on the oc forum
 
A few comments

"At conditions outside functional operation condition limits....."

The above has been on every Intel data sheet since the first C2D and probably long before.

XS people in general seem to have a "more voltage is better so a LOT has to be a lot better" mentality I do not understand especially 6 months later when the "it don't OC like it used to" posts start. To be fair we have a few here too, but its your CPU, do what you want.

Heat has never been an issue with C2Ds they will just shutdown if they get too hot. It is probably not a good thing to run it so hot as to cause this, but before the thermal trip there is the famous Tjunction "I am hot" temp where the CPU asserts a PROCHOT# interrupt and in theory you should be able to run the CPU 24x7x365 one deg C from this value at rated voltage - forever, at stock voltages.

Absolute Max Vcc.
One would be wise to heed Intel's recommendation/spec in this area. If one takes the trouble to delve into how this number is arrived at you find that there is actually little destructive testing done. Intel is so familiar with their manufacturing process and the physics of the materials used they are very confident that in knowing exactly how the CPU is made, the materials used and the physics and dimensions of the transistor junctions, that they can very accurately, based on materials science and the physics, calculate at what max voltage the transistor junctions will fail pretty accurately. The downside is that there are always "exceptions" as this is an average calculation. Some CPUs can take more, some less, but in general over the entire population of CPU;s manufactured it is a very good number. The thing to note is that Intels "functional envelope" is way way below this value (VCC max) and thus the big disclaimer paragraphs. You are on your own. Again its your cpu and do what you want. You can cool it all you want but at some voltage the dielectric between the transistor juntions will fail and you will destroy the transistor gate and usually transform the CPU into a keychain.

One would also be wise to understand why Intel REQUIRES Vdrop/droop in the design of cpu power requlation circuity. In short, at high Vcore the "overshoot" of the voltage requlation circuit as it tries to maintain a stable Vcore as the load current radically changes (say firing up Orthos from an "idle" state for example) will put higher voltage spikes on the Vcc that can kill the cpu. A techincal discussion is here: http://www.thetechrepository.com/showthread.php?t=126 The reference link a the bottom to the Intel power design guide is good reading for anyone that intends to run at very high Vcore but cannot afford to buy new processors every month. If you can, well who cares then, go for it.
 
HarkovAdm said:
no one else has anything to say about this problem? It seems a lot of sites I go to talk about '4.0ghz+ with over 1.4v'

If they are volting that high, and some people are saying there are electromigration problems, I am confused why this is not the #1 topic on the oc forum
Click to expand...

It's odd that some chips require that much voltage. My chip (45A576) will do 4ghz with 1.26v and 4.5ghz with 1.40v.
 
-=Antimatter=- said:
What kind of temps are you getting with your Ultima-90? My e3110 is coming today and I'm still undecided on how far I want to clock it.
Click to expand...
I'm at work right now, but if memory serves me, I max out at 70C after running Otheros Small FFTs for 8 hours. Under normal load I dont really break 65C. Idels anywhere between 28C and 40C depending if I cold boot or let it cool down from load.
Ambient room temp is around 25C tops.
 
You must log in or register to reply here.
Back
Top