I plan to put a Claymore-like watchdog: in case a GPU falls under a certain hashrate (default: 0) the miner quits. Then it's up to you to have a script (.bat or .sh) whose next line is a reboot command.

The watchdog wouldn't apply during the warmup session, since then the hashrate is very unstable and inconsistent, on purpose. It may crate both false positive (hashrate higher than real hardware max) or false negative (hashrate very low). That's because the VRAM memory accesses are managed to ensure 100% hashrate after the warmup. Something that was broken in the 0.32+
And not applied to CPU, since a slower CPU is caused by a legitimate other app that need some compute (the system, a driver, whatever...).

0.32g released soon. I skip the minor updates and will release them later in a minor upgrade. 0.32g will only have:


Then next: