We observe a phenomenon on a custom board that is equipped with an i.mx7d and 512MB lpddr2.
What could be the reason for this behavior?
Thanks and best regards
Thanks for your comments.
1. Yes, we ran the DDR Stress Test on corcerned boards overnight. The test passed.
2. We haven't tried to reproduce this issue on the i.MX7 SABRE board, but we ran the md5sum test on 100 custom boards. Three boards failed.
Could you please share your thoughts on my above statements.
Is there any specific verification strategy/test that targets the processor's cache memory?