Update: I think I was able to get to the bottom of it. A while ago, after upgrading the server, I set a bunch of power efficiency options in BIOS to minimise the hardware impact on the power bill. However, it seems these settings may impact the system stability at high loads, which is what I've been experiencing.
I also played around with the EXPO profile a little bit since I've seen system instability with some stock profiles in the past.