Bendable non-silicon RISC-V microprocessor.

Area of Science:

• Materials Science
• Computer Engineering
• Electrical Engineering

Background:

• Semiconductors are crucial for scientific research and connectivity.
• Future hardware advancements will impact quantum computing, AI, and edge computing.
• Open hardware fosters collaboration in education, research, and industry.

Purpose of the Study:

• To present Flex-RV, an ultralow-cost, flexible 32-bit microprocessor.
• To integrate machine learning (ML) hardware acceleration into the microprocessor.
• To demonstrate new instructions for extending the RISC-V instruction set for ML workloads.

Main Methods:

• Fabrication of a 32-bit microprocessor using indium gallium zinc oxide thin-film transistors on a flexible polyimide substrate.
• Integration of a programmable ML hardware accelerator and custom RISC-V instructions.
• Testing functionality and performance under flat and bent conditions on a flexible printed circuit board.

Main Results:

• The Flex-RV microprocessor operates at 60 kHz with less than 6 mW power consumption.
• Demonstrated functionality under both flat and tight bending conditions with minimal performance variation (average 4.3%).
• Successful execution of programs on the flexible microprocessor, showcasing its bendable nature.

Conclusions:

• Flex-RV pioneers sub-dollar, open-standard, non-silicon 32-bit microprocessors.
• This development democratizes computing access and enables emerging applications.
• Potential applications include wearables, healthcare devices, and smart packaging.