Not sure if this is applicable for high-end audio, but nevertheless an interesting idea...

Audio Pixels: "An Audio Pixels speaker is a MEMS chip roughly 1mm thick. The chip replaces conventional speaker driver(s), enclosure or acoustic chamber, as well as the electronic circuitry associated with converting the digital signal feed to analog (via a D2A + Power Amplifier)."

The Audio Pixels speaker chip is not only radically smaller; it consumes a fraction of the power, to produce far better acoustic performance (clear sounding, less distorted, more accurate sound reproduction). The qualitative differences are reflected in better performance in a substantially smaller package:

• Wider frequency range
• Lower frequencies
• Immeasurable distortion
• Louder sound volume
• Less power consumption

Audio Pixels speakers also offer unprecedented flexibility. In sharp contrast to conventional speaker technologies where the types and number of drivers, as well as the enclosure and its electronics all must be customized and endlessly fine-tuned to a specific application; Audio Pixels offers a single chip design that is modular and perfectly predicable and linear. Given that the total number of “pixels” is the only factor determining quality, frequency and sound volume, manufacturers and device designers need only to determine the target acoustic performance and select the appropriate number of speaker chips, which are serialized using a single controller.

The homogeneous construction of the chips simplifies and reduces the cost of customer design, integration, assembly and product inventory. It also permits the Company to optimize its pixels, the array, the fabrication and test processes.