Hi Marty McSly,

The lack of CPU error codes doesn't necessarily mean the motherboard is fine. If CPU2 never even starts to initialize, the system may not get far enough to generate a CPU-related error. In other words, a missing error code simply means the system didn't reach the point where it could report a problem — it doesn't confirm that the socket, VRM, or board, are healthy.

 

Here are several steps worth trying before writing off the motherboard, beyond reseating or swapping the CPU:

1. Resolve the HWC2003 first.
   This code indicates a storage cable that is either disconnected or improperly connected. While it's unrelated to CPU2, it could obscure other entries in the SEL or iDRAC logs. Re-seat any storage, SAS, or backplane cables you may have disturbed while tracing the CPU2 power cable. After that, reboot and review the SEL for any CPU-specific entries such as CPU0000, CPU0010, CPU0023, CPU0702, CPU0703, or CPU0704.
   https://www.dell.com/support/manuals/en-us/poweredge-t420/t420ownersmanual/system-error-messages?guid=guid-c66b89ab-dc55-4faf-988c-1afb3fe72929&lang=en-us

 

2. Test each E5-2470 v2 individually in socket 2 with socket 1 left empty.
   Since both processors already work in socket 1, if one of them fails to POST in socket 2 while an E5-2420 succeeds there, the issue likely points to a problematic CPU rather than a faulty board.

 

3. Clear NVRAM/CMOS.
   The board may still be holding outdated configuration data from a single-CPU setup, including cached SPD, TDP, or P-state information. Use the NVRAM clear jumper on the motherboard, or remove the coin-cell battery for a few minutes, then boot with both CPUs installed using default BIOS settings.

 

4. Boot with a minimal hardware configuration.
• Install just one DIMM per CPU in the correct first white slot for each — A1 and B1 on the T420.
• Remove all PCIe cards, drives, and backplane connections.
• Run from a single PSU only.
• Boot into the BIOS, iDRAC, or run the F10 ePSA diagnostic and check whether CPU2 appears. This helps eliminate the possibility of a misbehaving expansion card or power distribution issue causing the problem.

 

5. Carefully inspect socket 2.
   Even though an E5-2420 works in that socket, the E5-2470 v2 may place a slightly different load on certain pins. A bent, compressed, or contaminated pin could cause an intermittent contact that only shows up with one CPU model and not another. Look closely for any thermal paste residue in the socket and verify that both the retention bracket and heatsink are fully and evenly seated.

 

6. Try a different LGA1356 CPU in socket 2.
   You already know the E5-2420 is recognized there. If a completely different E5-2470 v2 sample works in socket 2, then one of your original CPUs is likely defective. However, if every E5-2470 v2 you try fails in socket 2 while an E5-2420 succeeds, the CPU2 VRM or QPI path on the board may be marginal — capable of handling the lower-core processor but unable to sustain the higher-core part.

Bottom line: If every CPU tests fine in socket 1 and only the E5-2470 v2 pair refuses to work in socket 2, the motherboard — specifically the CPU2 socket, VRM, or QPI circuitry — becomes the most likely remaining cause. The fact that an E5-2420 works in that socket keeps the board in the conversation as a partial success, but it doesn't rule out a board-level fault that only surfaces under the demands of the higher-core processor.

 

Regards,

DELL-Bridjet V

Social Media and Communities Professional

Dell Technologies | Enterprise Support Services

#IWork4Dell

Was your query answered? Please click Thumbs Up on the post that helped to Mark it as Accepted Answer!