Hybrid Digital-Neuromorphic Computing Services: Interoperability, Interfacing and Programming

Purpose and goal

This project focuses on the challenges associated with the integration of Neuromorphic Computing Systems (NCS) into existing digital infrastructure. Specifically, it addresses how brain-inspired sensor systems and processors that use event-driven, asynchronous, and dynamic components for resource-efficient information processing and machine learning can be more easily programmed and integrated from cloud to edge. The goal is to enable long-term sustainable and massively distributed artificial intelligence (AI).

Expected results and effects

The project is expected to promote industrial renewal, strengthen the national skills supply, and contribute to Sweden´s technological resilience by establishing proof of concept in the lab. Specifically, digital interfaces to NCS will be developed, facilitating interoperability and declarative programming of NCS based on open standards and optimization through learning, addressing the growing demand for energy-efficient and robust AI from network edge to cloud.

Planned approach and implementation

This project will use an iterative, agile methodology to achieve the goals. During the initial phase we will develop a digital interface template and declarative programming concept. The subsequent phase will focus on creating data and communication abstraction concepts. Each phase will have regular project meetings to ensure alignment of activities. Relevant actors and potential industrial users will be involved through feedback sessions and a workshop to ensure that the project´s conceptual solutions are applicable and beneficial to the end-users.