Parallel mechanical computing: Metamaterials that can multitask

Area of Science:

• Physics
• Materials Science
• Computer Engineering

Background:

• Analogue computing, particularly wave-based systems using metamaterials, is regaining interest for direct input processing.
• Current analogue computers are limited to single-task operations, hindering broader computational applications.
• The inability to perform multiple tasks concurrently limits the advancement of mechanical computing devices.

Purpose of the Study:

• To present a novel pathway for analogue mechanical computers to process independent computational tasks simultaneously.
• To overcome the single-task limitation of current wave-based analogue computing systems.
• To enable multitasking capabilities within a single architected structure.

Main Methods:

• Utilizing metasurface building blocks with broken time invariance.
• Generating multiple frequency-shifted beams from a fundamental signal.
• Assigning distinct computational tasks to independent frequency channels.

Main Results:

• Demonstrated a method to simultaneously process independent computational tasks within the same architected structure.
• Achieved self-generation of multiple frequency-shifted beams by breaking time invariance.
• Showcased tunable harmonics enabling multitasking in analogue mechanical computing.

Conclusions:

• The developed metasurface approach allows analogue mechanical computers to multitask effectively.
• Breaking time invariance in metasurfaces is a viable strategy for parallel analogue computation.
• This advancement opens new possibilities for more capable and versatile mechanical computing devices.