Life expectancy of a 920 @ 4.2Ghz @ 1.45V?

M Powered · Feb 3, 2006

Right now my rig is pretty solid running at the speed in the topic.

Question is how long will this OC last my some say 6 months to 2 years, is this true?

iamjcl · Feb 3, 2006

...Im not sure anybody really knows, but I would like to get some informed opinions on this as well - my chip requires the same voltage to do 4.2 - really, I'd say maybe 1.4625 to have some safety margin, and this is at 47c, 12-hour loaded temps. So I'm running at 4.0, which only takes 1.375.

- Chris

crimedog · Feb 4, 2006

longer then you'll keep the chip

3DFlyer · Feb 4, 2006

I'd also say your fairly safe, but that's no gurantee. I'm gonna be watching the 1.55 and up CPU's though. There was some discussion if these 65nm chips may not like that 1.55v and higher over the long haul. I belive it was proth that was gonna do some temperature testing, but I can't remember what thread it was in or if he posted about the tests. I belive he was concerned about temperature spikes...spikes that happen quickly and have to be tested with special equipment becasue the mobo software doesn't sample fast enough. The theory was that in between the sampling rate temp spikes could occur and do damage.

If anybody has any further discussion on that it would be most interesting. If there are differences in the 65nm chips from a OC'ing standpoint I would like to hear about it. Anything voltage related is always a welcome sight!

M Powered · Feb 4, 2006

3DFlyer said:
I'd also say your fairly safe, but that's no gurantee. I'm gonna be watching the 1.55 and up CPU's though. There was some discussion if these 65nm chips may not like that 1.55v and higher over the long haul. I belive it was proth that was gonna do some temperature testing, but I can't remember what thread it was in or if he posted about the tests. I belive he was concerned about temperature spikes...spikes that happen quickly and have to be tested with special equipment becasue the mobo software doesn't sample fast enough. The theory was that in between the sampling rate temp spikes could occur and do damage.

If anybody has any further discussion on that it would be most interesting. If there are differences in the 65nm chips from a OC'ing standpoint I would like to hear about it. Anything voltage related is always a welcome sight!

Please don't tell me about temperature spikes, I just got this water system fully setup less than a month ago - I don't need any temptation to go Phase change.

3DFlyer · Feb 4, 2006

M Powered said:
Please don't tell me about temperature spikes, I just got this water system fully setup less than a month ago - I don't need any temptation to go Phase change.

hahaha Phase...you know you want it...it's hauting you...you will submit.

I'm messin with ya... I know you've spent piles on hardware, so if you have a wife, please don't! LOL

Ross · Feb 4, 2006

You don't need any specialized equip...if you're on air in particular or water, open SpeedFan, click Charts->Temp->CPU. Presler/CM are spastic compared to all of my Prescotts, which were almost always a flatline for idle temps.

proth · Feb 4, 2006

3DFlyer said:
I belive it was proth that was gonna do some temperature testing, but I can't remember what thread it was in or if he posted about the tests. I belive he was concerned about temperature spikes...spikes that happen quickly and have to be tested with special equipment becasue the mobo software doesn't sample fast enough. The theory was that in between the sampling rate temp spikes could occur and do damage.

I was concerned about overshoot, lets say you have Vcore @ 1.55 and then it droops to 1.45 by running SP1M. When SP stops there could have been a rapid spike up to 1.6 or higher before settling back down th 1.55. These rapid overshoots could degrade the CPU after time.

Anyway, there are no spikes. Vcore simply droops down in a few milliseconds and then droops back up in a few milliseconds.

As for my opinion, I think both high Vcore AND temps can limit CPU life to less than a year. But this is probably running it past the manufacturer's limits of 1.6 and 60c for extended periods. Any Presler or Cedar Mills should last "a long time" if constant temps are under 60 and voltage is under 1.6.

iamjcl · Feb 4, 2006

Proth,

1.6 is intels spec for the preslers? I know 1.3 is the max that they state "on the record', so I suppose you've found a white paper or something saying its been engineered with a 1.6v limit in mind ?

- Chris

proth · Feb 4, 2006

iamjcl said:
Proth,

1.6 is intels spec for the preslers? I know 1.3 is the max that they state "on the record', so I suppose you've found a white paper or something saying its been engineered with a 1.6v limit in mind ?

- Chris

Datasheet for 9xx has max Vcc and FSB voltages @ 1.55.
Digging through the white papers shows reliability testing was done @ 1.6
This thingy may hang if you run it @ 1.3
ftp://download.intel.com/design/PentiumXE/datashts/31030602.pdf

M Powered · Feb 4, 2006

So 4.2 Ghz @ 1.465 will last a long time?

proth · Feb 4, 2006

M Powered said:
So 4.2 Ghz @ 1.465 will last a long time?

quite a few years if not constantly run above 60c.
IMO

What are your idle and load temps.?
Have you ever checked max OC at 1.55. ?

josh478 · Feb 4, 2006

how hard would you guys say it is to do the vdroop mod? im seriously considering it, because i know i can get a hell of a lot more voltage out of my chip with it... it drops almost a complete .1v under load.

like, i have it set to 1.425 in bios, at idle its at 1.48 then at load its at 1.39 and fluctuates to 1.40

d94 · Feb 4, 2006

i run 4.4 @ 1.45..if anything mine will "deteriorate" faster lol
but i have a feeling ill have a different chip 6 months from now

josh478 · Feb 4, 2006

yeah, ill probably ebay my chip when conroe comes out. i just hope my board will support a conroe.

Ad Rock · Feb 4, 2006

proth said:
As for my opinion, I think both high Vcore AND temps can limit CPU life to less than a year.

I was just wondering what you are basing this on?

Do you have any numbers or links that we could look at?

I would just like to take a look at the information you are using to make this statement.

isp · Feb 4, 2006

while you're at it, could you please give me tomorrow's lotto numbers...

M Powered · Feb 4, 2006

I've already reached my goal of 4.2Ghz, I don't see the need to go any further, especially since I have a VGA waterblock installed - which already generates enough heat.

Under load 12+ hours of operation max temp is 48C Idles at 32-34C/36

josh478 · Feb 4, 2006

my max temps are 48c with 1.4v at 3.6 going to keep lowering the voltage until i cant get it to boot.

down to 1.38v stable with pcmark04 71xx score

M Powered · Feb 4, 2006

josh478 said:
my max temps are 48c with 1.4v at 3.6 going to keep lowering the voltage until i cant get it to boot.

down to 1.38v stable with pcmark04 71xx score

Getting it to boot is not a very good sign of stability as you probably already know.

I could probably get my 4.2 to boot at a much lower voltage, but when I open something its probably going to crash. Right now I am 1 knotch above my "stable" setting just for good measures.

Life expectancy of a 920 @ 4.2Ghz @ 1.45V?

Member

Member

Member

Banned

Member

Banned

Senior Intel-lectual

Member

Member

Member

Member

Member

Member

$30 a phone

Member

Member

Registered

Member

Member

Member

Similar threads