"Burnning in"?

T

Tony Montana

[H]ard|Gawd
Joined
Sep 11, 2001
Messages
1,336
Has anyone ever heard this word? It has been explained to me that you need to burn in a CPU at a certain speed before OCing more? Is this correct or can you CO right to the speed you want?
 
A while back I had a Celeron 433 that ran at 590(90FSB). When I first was finding the max overclock, the system would not post at anything higher than 590. After running it for a few months at 590, I was able to use boot into windows and do some benchmarks at 620(95FSB), and occasionally post at 650(100FSB).

That is the only processor I owned that seemed to "burn in".
 
Some people might say that burning in is just a myth... its debatable.
 
Agreed.

I've had my 2.4c for over a year now and running it at 3.4 hasn't magically unlocked any higher speeds.
 
when someone one says something about burning in and computers the first thing that somes to my mind is the time it takes artic silver to properly set up.
 
MooCow said:
Some people might say that burning in is just a myth... its debatable.
Click to expand...
Funny... I always thought that the main purpose of burning in was to test the processor's stability. The logic behind it (as I learned it... feel free to correct me, though, if I'm wrong) is that the burning in process (often done through a benchmarking/diagnostic program such as Sandra and Prime95) was simulating the CPU at full load for at least a few hours, and that if the CPU could handle the stress caused by the burn-in, then it should handle nearly anything else. It was often done after the processor was overclocked to make sure that the CPU could handle a full load in its new settings.

Since I don't overclock, I can't call on the validity of burning in after an overclock, but I've always burned in my processor after the intial build of each rig that I built. I haven't had any CPU-related problems in any of the rigs that I built, but that may be more attributed to luck than anything else....
 
Burn-in/Stress test is what I do to every part that I get new.

With electronics, if they survive a stress-test for a few days, they more than likely will survive for a very long time. Electronics tend to crap out real quick if they are prone to in the first place.
 
tiraides said:
Funny... I always thought that the main purpose of burning in was to test the processor's stability. The logic behind it (as I learned it... feel free to correct me, though, if I'm wrong) is that the burning in process (often done through a benchmarking/diagnostic program such as Sandra and Prime95) was simulating the CPU at full load for at least a few hours, and that if the CPU could handle the stress caused by the burn-in, then it should handle nearly anything else. It was often done after the processor was overclocked to make sure that the CPU could handle a full load in its new settings.

Since I don't overclock, I can't call on the validity of burning in after an overclock, but I've always burned in my processor after the intial build of each rig that I built. I haven't had any CPU-related problems in any of the rigs that I built, but that may be more attributed to luck than anything else....
Click to expand...
I believe running Prime95 or Sandra is to test the stability of the system... burning in is rather making it run hot in hopes of boosting performance. I think... now you're all confusing me.
 
Burning in is much like letting the thermal compound adjust to the average heat so it can efficiently work at its best. While Prime 95 and other programs does its main job well (stability), it also helps with burning in. Because Prime 95 puts extra heats on the CPU, the thermal compound and the electronics will adjust to that temperature. In the process of all this, everything adjusts to work efficiently. Burning in is pretty much like that. Burning in also allows increases to headroom so this is where overclocking comes in.

-J.
 
You must log in or register to reply here.
Back
Top