• Notice regarding image hosts and attachments. See here.

Sandy bridge break in period?

Justintime

Justintime

Justintime

Soldato
Joined
24 Jul 2006
Posts
8,876
Location
Hoddesdon, London, UK
Have seen this mentioned in a few places and have 4 tales very similar, thought it maybe was a just on the line overclock but maybe not?

2500K - originally good for 4.6Ghz 1.35v - now needs 1.365v to be 24hr prime and 100 run LINX stable, constant BSOD 124 after 15 weeks at original vcore

2600K - 4.8GHz 1.375v , now needs 1.39v 17 weeks

2600K - 4.6Ghz 1.34v , now needs 1.35v 6 weeks

2600K - 4.5GHz 1.36v - poor clocker and now needs 1.365v after 4 weeks

However, i've built about 30 for the year with about 1/4 being overclocked and the rest report no problems but i'm not there to run prime etc.. to check if they're truly stable.. anyone else notice this?
 
Last edited:
What do you meen by break in period? Its not exactly a break in period if you need more voltage for the same overclock as before.Thats signs of degradation no?
 
I'm reading that since most people have passed this stage it generally stops there and does not degrade further. I noticed this on a few Bloomfield chips, the 920 i have here at 4Ghz 1.33v originally did that at 1.3v for about 4 weeks then would BSOD almost instantly on Windows bootup, however its being used 24/7 and a year later is still prime and LINX/IBT stable
 
Surveyor said:
I've had my 2600K since mid January on around 1.34V for 4.6GHz.

Never had to change a thing.

PC on 24/7.
Click to expand...

I'm thinking a strong overclock with enough vcore headroom might not show slight degradation, which i think is the case here, these chips were probably on the edge of stability while the rest have enough headroom and may never show it.
 
It's difficult to think of a reason why a chip would run better after a period of time. The closest I can come is proposing that the weaker areas have been broken, and the remainder will hold together reasonably well.

Changing to 32nm while being enthusiastic about voltages was always going to lead to some chips degrading earlier than expected.
 
JonJ678 said:
It's difficult to think of a reason why a chip would run better after a period of time. The closest I can come is proposing that the weaker areas have been broken, and the remainder will hold together reasonably well.

Changing to 32nm while being enthusiastic about voltages was always going to lead to some chips degrading earlier than expected.
Click to expand...

Did you read my post? :D
 
Yes, but I don't believe the edge of stability is the same thing as the point of significant degradation.

Unless by vcore headroom you mean the unknowable voltage between that which is set and the point where it will degrade within an unspecified time period. Which I suppose you might do.

Or you're suggesting deliberately running at higher voltages than required, so that when it starts to degrade (sooner than necessary), it takes longer to notice. Which doesn't make much sense, since chips degrade continually at any voltage, merely at different rates. It's also inherently strange to deliberately overvolt relative to that which is required.

So I guess I read it but didn't understand. It's late after all.
 
I think he means by edge of stability is if you get a cpu stable at 1.35 it and you lower it to 1.3495 it will still pass test but say you ran a test for 24 hours rather than the 12 youdid with 1.35 then it failed it would be on the edge.

I always run my chips with a little more than i need.

I do 10 mins testing with Intel burn and prime then carry on as normal. If i .get a blue screen i bump the volts and carry on as normal till it happens again.

I run my chip at [email protected] 24/7 im happy with that.

I could probably lower it a little but can't be bothered testing.

So im not on the edge so I've not noticed op.s issue. But i have heard about it with older i7 chips.
 
Last edited:
Hi

I've just revved my 2600K up to the 'proper' as-purchased speed of 4.6 from stock 3.4 & after a few days it gave a 124 BSOD during 100% usage. Then after reset, just 3 hours. Trying now on 3.4 again. Do I need to up my voltages slightly for my particular example? All I did was up the multiplier.

Dave
 
Have my 2600k running at 4.5ghz @ 1.31vcore, low v droop, spread spectrum disabled. Been like this since 6th Jan, no issues to date. Have run upto 5.7ghz at 1.6vcore on suicide runs.
 
dave_beast said:
Have my 2600k running at 4.5ghz @ 1.31vcore, low v droop, spread spectrum disabled. Been like this since 6th Jan, no issues to date. Have run upto 5.7ghz at 1.6vcore on suicide runs.
Click to expand...

I run my 2500k at 4.7 at 1.325 volts 24/7 & have even pushed it up to 1.7 volts for o/clocking, best i got was 5.428 at 1.66 volts.

At 1.7 volts the bios was showing a temp reading of 97 degrees so i guess i won't try again until i get better cooling.

I don't believe in running Prime or any other benchmarking programes as my chip will never be at 100% for 8 hrs, or whatever is the normal rum for Prime, if it runs my normal programes 24/7 without BSODing then its stable for me.
 
I've been running my 2500k at 4.5GHz / ~1.31V since early January (one of the lucky ones who got SB before the official release). I spent some time then trying to find the minimum voltage required. It blue screened with prime with 0.005V less, so I know I'm just the right side of stable, but really close to the edge.
I haven't had to increase the voltage, still perfectly stable.
 
I think i'm chalking it up to on the edge clocks and possibly due to higher ambients etc.. since most of the clocks were done in cooler months. None of the revised clocks has shown any issue since.
 
HumberView said:
Hi

I've just revved my 2600K up to the 'proper' as-purchased speed of 4.6 from stock 3.4 & after a few days it gave a 124 BSOD during 100% usage. Then after reset, just 3 hours. Trying now on 3.4 again. Do I need to up my voltages slightly for my particular example? All I did was up the multiplier.

Dave
Click to expand...

Very likely, more like 1.3v + for 4.6Ghz.
 
Had my 2500k since 1st of march.

Had it at 4.8ghz @ 1.290v since day one.

Yes 1.290v , i must have got really really lucky with my chip, never had to change a thing, never a single crash, lag, hang, or anything

:)
 
You must log in or register to reply here.
Back
Top Bottom