Quick overview of different JUCE vectorisation options

Could somebody please outline the various options that JUCE now offers for audio processing vectorisation, and how they compare to each other?

What’s the difference between using floatVectorOperations and SIMDRegister?