DL740 G1 Accelerator Cache Memory Parity Error Socket 1

[MT19]

Hi -

We have a DL740 G1 server and when we power it up, it crashes during the memory count with a 'Accelerator Cache Memory Parity Error Socket 1' error message. I've broken the configurations down to having slots 1 & 2 populated with 512MB DIMMS on all five of the memory cartridges. Also, these are not the original DIMMS that were populating slots 1 & 2, assuming that maybe one of the original DIMMS were bad. We are, however, still getting the same exact error indicating the same slot #. We are unable to do any firmware updates, since we can't get past the memory count in POST.

I have ordered a new memory cartridge and a system board. Please let me know if I am headed in the right direction with ordering those parts and if anyone else has any other recommendations.
Thanks much,
mark

[Brian_Murdoch · Accepted solution]

Hi Mark,

The Accelerator cache referred to is for the processors. The actual cache is located on the system board so you may have a problem here or on the Processor card itself. Some servers (8500, DL760 for example) generate a 220 error when the processor cache fails and you can go into the BIOS and "Set Cache Accelerator Corrected" but it seems you can't get into the BIOS in any case.

Try the system board and let us know how you get on. If it is possible to get into the BIOS before that you may be able to clear the cache problems but it sounds like it is more serious than simply resetting it.

Regards,

Brian

[MT19]

Thanks Brian, you were right. It was the system board. I tried replacing some of the memory cartidges first and breaking down/swapping the memory DIMMS around, but no luck there. Once I replaced the system board everything booted up perfectly.
Thx again,
mark