In an industrial environment where large induction motors are control by VFD, the high frequency generated by the VFD (harmonic distortion) induce a high voltage over the bearing and cause the bearing pitting and ultimately to fail. The known solutions to such problem are:
1. use motor with a ceramic ball bearing.
2. Grounding brushing on the motor bearing body to shunt the induced voltage to shaft.
Background information if bearing noise is to be blamed.
For typical industrial motor (for north america), the fundamental frequency is 60Hz. The motor speed is a function of the frequency applied, varying the fundamental frequency from 0 to 60Hz changes the motor speed. In generating the varying frequency by VFD, AC is first converted to DC by converter, and then the DC bus voltage is pulse width modulated to produce the varying frequency by an inverter. The inverter chops the DC voltage to produce the fundamental frequency but also generates harmonic frequency of several orders, 3rd to 50th. The harmonic frequency is the culprit in this induced voltage across the bearing if not properly grounded.
Tesla VFD DC bus voltage is supplied by the car battery, therefore no need for the AC-DC converter, but the DC-AC inverter is required. Tesla drive motor runs at tens of thousand RPM, much higher than the typical 1800RPM of industrial motors, therefore the fundamental frequency of the VFD is already much higher than 60Hz. I don't have the exact number, but the theory is valid. This higher frequency VFD also generates higher harmonics, aggravating the bearing arcing problem.
Others claim that the whinning noise is from the gears, please refer to their post for further insight. The exact cause of the noise is yet to be determined.