Safe vCore on P4 Northwood

Aces Hardware claims here

http://www.aceshardware.com/read.jsp?id=45000288

that using a vCore above 1.65v on a Northwood is not safe. They say that at 1.7v a Northwood CPU will only last a few weeks

According to them, the .13micron process used on the Northwood might allow electromigration in the transistors when higher voltages are used.

Anybody hear anything that would confirm or dispute this?

I don't know weather to believe them or not-- I'm running my 1.8A at 1.75vCore-- Ive got a great 43% overclock at this voltage and I'd really hate to turn it down but it'd be total crap if the CPU fried itself by the middle of March

i'm running at 1.85v the 1.8A, they're just talking crap

The question, dude, is how long will it run at high voltage--

I do think he's most likely got a load on about this

It seems to me that upping the voltage by 15% to 20% over stock isn't going to shorten the life of your CPU from years to weeks but I'd really hate to be wrong

All the same, I don't think I'll back mine down unless someone can shed some more light on this

and the p4 .13 micron voltage warnings MIGHT apply to tualatin cpus too... if any1 got some reliable information, not misinformation, plz post.

If your chip is stable and cool at 1.8v or higher then I would not worry about it. I have seen some hit 2.1v with these northwoods.

Don't worry about ruining your chip too fast.

Pinky has been running his machine with 1.8v or higher for awhile now and I think he is doing okay. My chip is not as stable at a higher vcore than 1.65v usually.

I believe anything under 1.7 is fairly safe. I wouldn't want to push much beyond 1.7. I think anything over 1.8 is really asking for an early retirement, especially without good cooling. Some of these people using mods to get 2+ volts are really asking for it imo.

Are the default vcore like 1.475?

TC said:

I believe anything under 1.7 is fairly safe. I wouldn't want to push much beyond 1.7. I think anything over 1.8 is really asking for an early retirement, especially without good cooling. Some of these people using mods to get 2+ volts are really asking for it imo.

Click to expand...

Nah, 1.8v is okay if it can do it. You have to remember these chips have copper interconnects. Copper has a much higher melting point than Aluminum which the .18 cumine is on. I would not worry about 1.8v if the chip can handle it. Some can handle the higher vcores and some can't. The ones that can not handle the higher vcores usually clock higher at lower temps.

Hilight said:

The question, dude, is how long will it run at high voltage--

I do think he's most likely got a load on about this

It seems to me that upping the voltage by 15% to 20% over stock isn't going to shorten the life of your CPU from years to weeks but I'd really hate to be wrong

All the same, I don't think I'll back mine down unless someone can shed some more light on this

Click to expand...

Mine is running @ 1.8 ~ 1.825 playing back and forth with it for the past 2 weeks.. Nothing is going banana here, I think what they say is trash

I downloaded the P4 Spec sheet from Intel last night. It lists 1.75v as the maximum stress rating and says that 'extended exposure to the maximum rating may affect device reliability'.

Now if we translate the Intel-speak this would mean that at 1.75v the chip might not last as long as it does at stock--- but we all know already that bumping the voltage might shorten the life of a CPU from 10-15 years to 5-8 or so, but no one is gonna be running their P4 in five years anyway.


If Intel thinks 1.75v 'may affect reliabialty' then we're probably good to go as high as 1.85v and still get 5 years out of your processor.

Aces Hardware has definitely got a load on


BTW ol' man-- this has nothing to do about the melting point of copper and aluminum-- we're talking about electomigration