Specify the largest floating-point data type that can be atomically loaded and stored. If Double is specified, Float will be considered atomic as well.