I've noticed this since 1701. I judge my overclock with 2 hours of OCCT. I start with settings which pass that test, and then decrease voltages until it fails.What's odd to me is that frequently, when I increase my voltage back to what had previous...