Near-data processing moves computation closer to stored information, reducing unnecessary transfers between storage and host systems. The nKV demonstration shows how this approach can accelerate key-value lookups, scans, and graph analysis on computational storage hardware.