The Promise and Challenges of Brain-Inspired Hardware
The promise of AI to transform society for the better has been promoted by tech-companies, scientists, and policymakers since the early 2010s. AI is considered a key technology for future economic growth and a driving force for achieving the United Nations Sustainable Development Goals. AI systems can help reduce resource consumption, produce less waste, and optimize energy efficiency in production processes. They can also monitor and predict environmental changes to support decision-making in precision agriculture and ecosystem management.
However, these promising narratives are accompanied by a growing critical discourse on the ethical, material, and political challenges that AI poses. Ethical concerns may refer to the transparency and reliability of AI, as well as the risk of reinforcing existing discrimination through algorithmic bias. The big data requirements for training AI models also raise data protection and privacy considerations.
In response to these concerns, the European Commission aims to build a regulatory framework for “trustworthy AI” that protects personal data, privacy, and non-discrimination. Additionally, the materiality of AI is gaining attention, as high-performance AI applications consume significant computing power and energy, leaving a considerable carbon footprint. The production of electronic devices on which AI runs also consumes a lot of energy and raw materials, incurring environmental costs.
The material backbone of AI, particularly the production of semiconductor chips, also raises political concerns. Global manufacturers of semiconductor chips are mainly located in Asia and the USA, making Europe’s access to computer chips vulnerable to supply chain disruptions. To address these challenges, the European Chips Act aims to increase Europe’s share of global chip production capacity to 20% by providing public investment in regional chip design and production.
Neuromorphic Computing: A Potential Solution
In line with European efforts to address the ethical, material, and political challenges of AI, the German Ministry of Research and Education (BMBF) funds the NeuroSys Cluster4Future. NeuroSys is a high-tech innovation cluster that seeks to build an innovation ecosystem around the development of neuromorphic computing hardware for AI applications in the Aachen region of Germany.
Neuromorphic computing denotes a computer chip architecture that emulates the neural network of the human brain. This chip architecture is expected to be more energy-efficient than computer hardware based on graphic processing units (GPUs), which are commonly used for training AI models. Not only does energy-efficient neuromorphic hardware promise to reduce AI’s carbon footprint, it can also foster data security and privacy because it can be used for mobile edge-computing devices, like sensors and smart watches, that process data locally rather than sending them to cloud services.
To develop neuromorphic computing hardware in tandem with AI applications that respect data protection and privacy concerns, NeuroSys bundles expertise from scientists, engineers, social scholars, industrial professionals, and municipal actors in an emerging innovation ecosystem. The goal is to build a semiconductor chip plant in the Aachen region that will produce neuromorphic computing hardware tailored to specific AI applications for autonomous driving, personalized healthcare, smart cities, the Internet of Things, and digitalization.
The Anatomy of the NeuroSys Innovation Ecosystem
The NeuroSys Cluster4Future consists of five projects, each focusing on a different research topic related to neuromorphic computing:
Projects A and B: Basic Research
– Project A studies the characteristics of memristors, material components that can change their resistance and are suitable for representing the weights between neurons in an artificial neural network.
– Project B examines optical signal transmission in neuromorphic hardware, using light to encode and transmit information instead of electronics.
Project C: Hardware-Software Co-Design
– This project brings together expertise from hardware-related circuit design, automated system design, and neuroscience to develop innovative circuit architectures based on novel devices and material systems.
Project D: Software Development and Use Cases
– Project D investigates use cases of neuromorphic hardware, preparing and optimizing software from specific application areas for neuromorphic hardware.
Project E: Societal Dimensions
– In addition to the technical projects, Project E examines the economically viable, ethically robust, socially desirable, and environmentally sustainable development of the innovations emerging from the other projects.
The collaboration between these projects follows a co-design approach, where the development of hardware and software is closely intertwined. Specific needs of algorithms and applications can influence the development of novel devices and material systems, while new algorithms and learning models are developed to exploit the technical capabilities of neuromorphic hardware.
This push-pull dynamic requires hardware and software developers to engage in a continuous collaborative process of alignment, for instance, between the compute complexity required for highly-performant applications and the capabilities of neuromorphic hardware.
Driving the Market Entrance of Neuromorphic Computing
The economic dimension of NeuroSys supports the cluster’s main objective: achieving technological sovereignty for Europe in semiconductor and AI research, development, and production. This aligns with the European Chips Act, which aims to increase Europe’s share of global chip production capacity to 20%.
To ensure the long-term usability of the innovations from NeuroSys, the economic dimension will evaluate possible business models and map the value chains of neuromorphic chips and associated products. Identifying potential cost savings in production and use of the hardware is crucial for commercializing NeuroSys innovations and stimulating entrepreneurial activities.
While the exact market potential of neuromorphic computing is difficult to estimate, as the technology has not yet reached market maturity, NeuroSys researchers anticipate three main categories of applications:
-
Existing applications supplemented by neuromorphic hardware: Neuromorphic hardware can be used to enhance the performance and energy efficiency of existing AI applications.
-
Applications that shift production to neuromorphic hardware: Some applications may transition to neuromorphic hardware if it offers equal or better performance compared to conventional technology.
-
Novel applications enabled by neuromorphic computing: New applications may emerge that are only possible with the capabilities of neuromorphic hardware.
The market for AI hardware, including neuromorphic chips, is expected to grow significantly in the coming years, from $15.7 billion in 2022 to $70.9 billion by 2026. This growth, coupled with the European Chips Act’s investments, suggests that NeuroSys could have a significant impact on the high-tech sector and labor market in the Aachen region, attracting students, researchers, engineers, and professionals to work in neuromorphic computing.
Building a Responsible Innovation Ecosystem
The societal dimension of NeuroSys focuses on the social order that enables and supports the emergence of an innovation ecosystem around neuromorphic computing in the Aachen region. The cluster aims to transform the former coal mining area into an “innovation hub” by creating a collaborative ecosystem of universities, companies, start-ups, local authorities, and other stakeholders.
Applying the multi-level perspective from socio-technical transition studies, NeuroSys seeks to take advantage of the instability in existing socio-technical regimes (e.g., technological, science, and market regimes) to break through with its neuromorphic computing innovations. However, transitioning a new technology from a protected niche into the mainstream market requires careful management of stakeholder collaborations and the integration of societal considerations.
NeuroSys adopts a quadruple helix approach, involving actors from research, industry, government, and civil society, to facilitate knowledge sharing and build trust. This is particularly important in the current geopolitical climate of the semiconductor industry, where “technology theft” by foreign competitors is a concern.
Moreover, NeuroSys aims to build a responsible innovation ecosystem that integrates ethical, social, and environmental considerations into the research and development process. This includes addressing issues such as trust in AI, the impact of neuromorphic computing on human autonomy, and the sustainability of the technology’s production and use.
To cultivate this responsible approach, NeuroSys encourages reflexive engagement among its diverse partners, enabling them to step out of their disciplinary comfort zones and critically interrogate their assumptions and practices. This is crucial, as the societal dimension of the project is not limited to an “add-on” task force but rather seen as an integral part of the transdisciplinary research and development process.
Navigating the Environmental Implications
The environmental dimension of NeuroSys takes a critical stance towards techno-fix narratives, which cast neuromorphic computing as a straightforward solution to reducing the environmental impact of AI. Instead, the project adopts a holistic view, examining the complex relationships between technological innovation, economic factors, and environmental sustainability.
One key concern is the potential rebound effect, where efficiency improvements in neuromorphic computing lead to increased energy demand and environmental costs. For example, if neuromorphic hardware is used to enable new energy-intensive AI applications rather than making existing products more efficient, the potential environmental benefits could be offset.
To address these challenges, NeuroSys incorporates reflexive exercises into the research and development process, encouraging actors to continuously consider the environmental aspects of their work. This includes probing the social, ethical, and economic trade-offs involved in the development and application of neuromorphic computing technologies.
By taking a critical, interdisciplinary approach, NeuroSys aims to move beyond simplistic techno-fix narratives and develop a deeper understanding of the complex socio-technical dynamics that shape the sustainability of neuromorphic computing. This will help ensure that the innovation ecosystem emerging around this technology is not only competitive but also responsible and environmentally benign.
Conclusion: Transformation Research in Action
The NeuroSys Cluster4Future exemplifies the model of transformation research introduced in this edited volume, where technological, economic, societal, and environmental dimensions of change are inextricably linked. By bringing this model from theory into practice, NeuroSys aims to reveal the opportunities and challenges that emerge in the research process, contributing to discourses on transformation research, Responsible Research and Innovation, and adjacent fields.
As the innovation ecosystem evolves, NeuroSys will study and interrogate assumptions about “innovation” and “ecosystem” concepts, strengthening attempts to adopt a holistic, reflexive approach to transformation research. Through this endeavor, NeuroSys seeks to build an innovation ecosystem around neuromorphic computing that is not only technologically advanced but also socially responsible, environmentally sustainable, and aligned with European values.
