I should write this down once-and-for-all.
Double Precision Is Not Needed for Many-Body Calculations: New Conventional Wisdom
Leveraging the bfloat16 Artificial Intelligence Datatype For Higher-Precision Computations
Double-precision FPUs in High-Performance Computing: an Embarrassment of Riches?