About - NeuroWare

Advancing collaborative research in brain-inspired computing

The NeuroWare Innovation and Knowledge Centre (IKC) is a multi-institution centre set up to accelerate the development of brain-inspired systems for next generation computing. It brings together a diverse team of academic researchers with a network of industry partners. It is an inclusive centre, open to further partnerships with a mission to build a better and more sustainable future for computing.

Collaborative Research

Multi-institution partnerships

Building a better future

Our research workstreams

The IKC’s research programme spans the full landscape of neuromorphic hardware, from established complementary metal–oxide–semiconductor (CMOS) approaches to emerging materials, photonics and biologically inspired computing. Four complementary Workstreams collaborate closely to turn breakthroughs into scalable technologies.

Workstream 1

Neuromorphic systems using established technologies

We are advancing neuromorphic systems built on today’s most mature platforms, including CMOS-based architectures, analogue/digital design and the next-generation SpiNNaker2 platform. The team will focus on issues such as:

Running real-world applications efficiently on neuromorphic hardware.
Tackling scalability and interconnect limits in large systems.
Developing neuromorphic vision and edge AI capabilities (AI running on-device, close to data).
Advancing analogue and mixed-signal neuromorphic computing and exploring commercial pathways for emerging architectures.

Workstream 2

Next-generation neuromorphic technologies

We are developing neuromorphic hardware based on emerging devices and materials, including memristors (resistive memory devices that combine storage and computation), analogue computing elements and two-dimensional (2D) materials. We aim to build highly efficient, error-tolerant and scalable hardware by:

Integrating memristive devices with analogue/mixed-signal CMOS circuits.
Improving device performance and reducing variability.
Prototyping systems that significantly improve energy efficiency and throughput over conventional approaches.
Exploring probabilistic and on-chip learning methods enabled by device variability and new machine learning approaches tailored to analogue hardware.

Workstream 3

Neuromorphic photonics

We are combining neuromorphic electronics with integrated photonics to create mixed-mode computing that uses the speed and parallelism of light with the efficiency and flexibility of electronics. We are:

Developing photonic-electronic prototypes that deliver ultra-fast, ultra-low-latency and energy-efficient processing.
Creating architectures well suited to future AI, edge computing and real-time decision-making.
Exploring sustainable fabrication approaches and new functionalities enabled by optical spikes, optical synapses and co-design with other Workstreams.

Workstream 4 (Underpinning)

Computational neuroscience

We are applying insights from biological intelligence to design energy-efficient, fault-tolerant hardware capable of real-time, lifelong learning. We are:

Drawing selective inspiration from biological brain systems – including sparsity, memory replay and adaptive synapses - to overcome challenges like catastrophic forgetting (when systems lose previously learned skills while learning new ones).
Developing hardware-specific algorithms and learning methods that exploit the parallelism and unique capabilities of neuromorphic architectures.
Working closely with all Workstreams to create robust, adaptive and scalable neuromorphic systems.