"breaking in" a new cpu before overclocking

Jul 10, 2007
12,050
3
0
i just built a new P43/e7200 rig and i plan to do a mild overclock to ~3GHz, but will just break in the cpu with some normal usage for a couple of days to a week before i kick up the FSB to 333 (planning stock cooling and stock voltage).

or do u guys just oc the crap outta the cpu right out of the box?
 

Idontcare

Elite Member
Oct 10, 1999
21,118
58
91
What technical purpose does "breaking in" a CPU serve?

Break in periods typically apply to things that have dimensional tolerances which are expected to initially change rapidly during use and eventually the rate of change in the dimensions flatten out.

I.e. mechanical components in a system (bearings, bushings, etc).

I can see an argument for "breaking in" a thermal paste that requires a such a period (AS5 for example) but not for a CPU, entirely needless.
 

Rike

Platinum Member
Oct 14, 2004
2,614
2
81
When I build a new rig, I'll prime for at least 12 hours, usually 24, as an easy check to see if anything will fail right off the bat. I don't know that's breaking in, but if something is going to go wrong right away, I'd rather have it go wrong while I still have the parts on my bench.
 

bharatwaja

Senior member
Dec 20, 2007
431
0
0
Originally posted by: Rike
When I build a new rig, I'll prime for at least 12 hours, usually 24, as an easy check to see if anything will fail right off the bat. I don't know that's breaking in, but if something is going to go wrong right away, I'd rather have it go wrong while I still have the parts on my bench.

+1 :thumbsup:

This is what I do too....
 

COPOHawk

Senior member
Mar 3, 2008
282
1
81
I dont do the Prime thing for that long...maybe 30 min max.

These are not automobile engines with moving parts :)

Frankly, if it POSTS, then I go straight for the Windows install, then Prime for 30 minutes and look for any errors, instabilities, etc. I do like to have the system powered on for at least a few hours straight to check for any potential problems. Then, I run Memtest to check for any memory errors before overclocking.

Usually when I have a bad part, it is typically bad right out of the box, or in the case of a few hard drives, it makes its appearance known after the full Windows error checking.

HTH...
 

Extelleron

Diamond Member
Dec 26, 2005
3,127
0
71
As Idontcare says, there is really no need to break in a CPU other than to break in the thermal paste. The chips are already put through intense burn-in tests after fabrication/assembly as they are tested for stability and binned, there is no need to repeat this at home. CPUs really don't fail at stock speeds/voltage; they are built to run for years and years on end without error. It is things like memory and hard drives that you need to worry about.
 

Drsignguy

Platinum Member
Mar 24, 2002
2,264
0
76
Originally posted by: Extelleron
As Idontcare says, there is really no need to break in a CPU other than to break in the thermal paste. The chips are already put through intense burn-in tests after fabrication/assembly as they are tested for stability and binned, there is no need to repeat this at home. CPUs really don't fail at stock speeds/voltage; they are built to run for years and years on end without error. It is things like memory and hard drives that you need to worry about.


I agree. All I do is install the chip and run it for about a day @ stock just to make sure no errors occur. After that, It's stock-no-more.:)

 

OCGuy

Lifer
Jul 12, 2000
27,227
36
91
I think I boot into Windows once, then I'm magically in BIOS about 5 seconds later.

 

DSF

Diamond Member
Oct 6, 2007
4,902
0
71
And breaking in a thermal paste like AS5 is a long process that doesn't make that much difference anyway.
 

aigomorla

CPU, Cases&Cooling Mod PC Gaming Mod Elite Member
Super Moderator
Sep 28, 2005
20,839
3,174
126
Originally posted by: Rike
When I build a new rig, I'll prime for at least 12 hours, usually 24, as an easy check to see if anything will fail right off the bat. I don't know that's breaking in, but if something is going to go wrong right away, I'd rather have it go wrong while I still have the parts on my bench.

thats called fault testing.

has nothing to do with breaking in.

As IDC said, the only possibly things in your computer that would need to "be warmed up" is the fans and hard drives.
 

Gillbot

Lifer
Jan 11, 2001
28,830
17
81
I shoot for at least a mild OC right out of the box and go from there. I don't think I even run a chip @ stock unless I absolutely have to. Not even for bios updates! :shocked:
 

mancunian

Senior member
May 19, 2006
404
0
0
The last 2, an A64 4200 and an E7200 have been overclocked right off the bat.

Only took the E7200 to 3.4 initially, then 3.8 the next day.

Idontcare explains it perfectly, the lack of moving parts means that a 'break in' period is completely unnecessary.


 

Ratman6161

Senior member
Mar 21, 2008
616
75
91
Originally posted by: aigomorla

thats called fault testing.

has nothing to do with breaking in.

As IDC said, the only possibly things in your computer that would need to "be warmed up" is the fans and hard drives.

Yes. Make sure you don't have any problems going on before you start overclocking

 

SlowSpyder

Lifer
Jan 12, 2005
17,305
1,001
126
Hehe... I'm not sure my Opteron 165 ever saw 1.8GHz (until my CMOS battery died anyway). But in my case I upgraded just the processor and knew all the other components were good.
 

Tweakin

Platinum Member
Feb 7, 2000
2,532
0
71
As it has already been stated, there really is no "breaking in" period for a chip. It works or it doesn't.

I always boot to defaults, Install a clean OS with all service packs and applicable drivers. I take screen shots for reference and start the oc process...
 
Nov 26, 2005
15,086
310
126
Get use to it at default speed. I usually spend about a week or two before I allow my self to even think of rolling the dice. I think I spent about a month or more with my 8400 before I tried 400fsb. To me, it's all about setting my mind at ease with it and controlling myself. Eddie Murphy once joked: if you starvin and someone hands you a cracker - you are like "god damn that was the best cracker i ever had in my life, that was a good cracker, what was that? a saltine!?!? no that was a ritz, that was a ritz cracker, thats what that was..." and then after you've had that cracker for a while, you're like "what do i got to eat? O' just some regular old crackas"
 

Cheex

Diamond Member
Jul 18, 2006
3,123
0
0
Originally posted by: Ocguy31
I think I boot into Windows once, then I'm magically in BIOS about 5 seconds later.

Now that's what I'm talking about. Although most times I boot into Windows with an overclock, then just go higher. :D
 

brencat

Platinum Member
Feb 26, 2007
2,170
3
76
After it's built and boots fine, I run memtest86 via floppy overnight to test RAM @ stock. After that, it's install the OS time, copy over some of my files and utilities, and then off to the races! So basically 1 or 2 days at stock tops before OCing.
 

Foxery

Golden Member
Jan 24, 2008
1,709
0
0
Originally posted by: Ocguy31
I think I boot into Windows once, then I'm magically in BIOS about 5 seconds later.

Slacker. I know in advance what to expect out of a new CPU, so I push it halfway there right off the bat, and there's no way I'm waiting several extra days to stress it the rest of the way. :)
 

Gillbot

Lifer
Jan 11, 2001
28,830
17
81
Originally posted by: Foxery
Originally posted by: Ocguy31
I think I boot into Windows once, then I'm magically in BIOS about 5 seconds later.

Slacker. I know in advance what to expect out of a new CPU, so I push it halfway there right off the bat, and there's no way I'm waiting several extra days to stress it the rest of the way. :)

I try for at least 1 bin higher on FSB right off the bat. If it's a 266fsb chip, I try for 333, if it's 333 I try for 400. My first windows load on a new chip always has at least SOME oc on it! ;)