Is it necessary to burn-in RAM for server-class systems?

Posted by ewwhite on Server Fault See other posts from Server Fault or by ewwhite
Published on 2013-06-25T03:05:39Z Indexed on 2013/06/25 4:22 UTC
Read the original article Hit count: 543

Filed under:
|
|

When using server-class systems with ECC RAM, is it necessary or even useful to burn-in the memory DIMMs prior to deployment?

I've encountered an environment where all server RAM is placed through a lengthy multi-day burn-in/stress-tesing process. This has delayed system deployments on occasion and adds an extra step to the hardware lead-time.

The server hardware is primarily Supermicro, so the RAM is sourced from a variety of vendors; not directly from the manufacturer like a Dell Poweredge or HP ProLiant.

Is this process useful? In my past experience, I simply used vendor RAM out of the box. Isn't that what the POST memory tests are for? I've encountered and responded to ECC errors long before a DIMM actually failed. The ECC thresholds were usually the trigger for warranty placement.

  • Do you burn your RAM in?
  • If so, what method do you use to perform the tests?
  • Has the burn-in process resulted in any additional platform stability?
  • Has it identified any pre-deployment problems?

© Server Fault or respective owner

Related posts about memory

Related posts about Hardware