I am about to assemble a 'puter with parts from some of the vendors at this site. Many of the SPCR posts I've read mention breaking in the CPU and heatsink. In the past, I've just assembled, flipped the switch, and started loading my OS. I have searched, but haven't found posts describing proper break-in procedures for components. Is there? What are they?

My rig will have:
-Case: Antec Solo (with rear case fan mesh removed)
-PSU: Seasonic S12-380
-Case Fan: Nexus 120
-MOBO: ABIT AB9 PRO P965 775
-CPU: INTEL C2D E4300
-HS: Scythe Ninja (passive)
-RAM: 1G G.Skill F2-6400CL4S
-Video: GeForce 7600GS
-HD: Raptor 150

Hello & welcome to SPCR,

I think what folks mean by this, is that some thermal pastes improve after a while, at normal operating temperatures. So, you don't have to do anything other than just use your computer, to "break it in".

BTW, the Ninja may or may not be the best choice -- depending on how it fits on that motherboard with that CPU. The mounting mechanism is the possible cause, so you may want to verify how it might work. The Ninja is one of the best HS, and certainly the best passive HS, but it has to mount firmly in order to work.

I'd suggest running Memtest86+ (http://www.memtest.org/) to make sure your memory is good. If it's not, you can get random crashes later and spend a lot of time and effort trying to work out what it is. You also may not be able to return it if you don't find problems early enough.

The only reason I'd suggest stressing the CPU etc., is to see what temps you get.

Thank you for your prompt response and advice. What about "burn-in"? I see a few posts mentioning CPUBurn. Is that to ID a faulty CPU?

Yes, that is good advice. I already have a bootable copy of that.

CPUBurn is a utility that stresses your CPU and will generate the highest temps you're ever likely to see on the processor. This can be good for diagnosing an incorrectly installed heatsink (e.g. your temps are well above what might be expected) or just configuring your fan speeds etc for the worst case scenario. Prime95 is also a great utility if you're trying to identify possible hardware issues (e.g. faulty cpu or ram).

"Burn in" can also refer to a long-standing myth in the overclocking community that took forever to die (and may still exist in fringe elements). The idea was, after a time of "burning in" a CPU could be clocked faster. The idea was wrong.

Memtest 86+ v1.70 at boot
Orthos when running Windows.

Orthos will load your cpu and ram.

Basically everyone covered it, there is no burn-in really other than the thermal interface material *possibly* working slightly better after some use.

However, stress-testing would be a good idea early on, to reveal any possible problems--stressing your cpu, ram, video card, power supply and overall cooling systems early on will help avoid problems down the road with unexplained crashes and the like.

I have gotten into the habit with my new builds now--especially when you are trying to do quiet systems with slow fans, you need to verify adequate cooling is in place.