Aspinity's RAMP™ (Reconfigurable Analog Modular Processor) technology is a neural-inspired processing platform that brings the versatility, performance, and consistency of digital systems into the more efficient analog domain.
The traditional neuromorphic computing concepts developed by Carver Mead in the 1980s used analog circuits to mimic neural systems. However, translating these large-scale analog neuromorphic concepts into deployable products has been stymied by challenges associated with low-power analog circuit design, such as offsets, PVT, and precise control of biasing for a large number of variables.
Aspinity has overcome these implementation hurdles by developing RAMP technology, the world’s first programmable analog neuromorphic processor. The RAMP platform leverages new and innovative analog processing techniques to finally actualize neuromorphic computing concepts, enabling a scalable analog processing platform that delivers:
- Complex decision-making capability: RAMP technology leverages non-linear analog circuitry to improve the performance of typical analog tasks, make decisions, and classify incoming sensor information.
- Ultra-low-power analog processing: RAMP technology is comprised of parallel, independent analog circuit blocks that operate in the subthreshold domain. Each of these blocks is implemented in a very small footprint and each can be independently powered only when they are needed for a specific task.
- Programmability: The complete functionality of a RAMP-based chip can be abstracted from hardware into software, enabling a flexible analog platform in which all aspects of the chip (connections of circuit blocks, parameters, etc.) can be programmed in software and stored in on-chip memory.
- Neural-Inspired Efficiency: Aspinity’s patented analog non-volatile memory (NVM) stores operating parameters and neural-network weights alongside the analog processing elements like real biological systems, resulting in an efficient, flexible platform that can process large amounts of data in real time.
- Repeatability and predictability: The programmability of RAMP-based chips allows software ‘trimming’ to offset the process variations of standard CMOS technology during high volume manufacturing, eliminating the challenges associated with implementing large-scale analog neuromorphic computing platforms.