Imagine a computer chip that works like a human brain instead of a traditional calculator. It processes information in quick bursts, which uses far less power and reacts much faster. This project shrinks that powerful technology from giant server rooms down to a small board that can fit into handheld devices.
Current AI hardware suffers from communication bottlenecks and excessive energy consumption. This prevents real-time, autonomous AI from running efficiently on small devices at the edge.
A single-chip PCB called the SpiNNode board and a software toolchain to deploy both deep and spiking neural networks on SpiNNaker2 hardware.
If you are a biotech company dealing with slow drug discovery processes — this project developed brain-inspired systems that accelerate optimization tasks. This allows for faster identification of molecular candidates while reducing energy costs.
If you are a drone maker dealing with high battery drain from AI processing — this project developed the SpiNNode edge board. It provides real-time, low-latency AI execution directly on the device to extend flight time.
If you are a factory operator dealing with communication bottlenecks in AI systems — this project developed a hybrid system for deep and spiking neural networks. This enables real-time autonomous control without relying on a distant cloud server.
Based on available project data, specific pricing and cost structures are not disclosed; the project focuses on creating a business plan for future commercialization.
Yes, the technology is scalable from a single chip up to supercomputer levels comprising 69,120 SpiNNaker2 chips at the largest server size.
The project explicitly aims to disseminate results while ensuring that the IP is protected during the engagement with external parties.
The project developed a deployment toolchain to map Deep Neural Networks (DNN) and Spiking Neural Networks (SNN) onto the hardware.
The project runs from 2023-05-01 to 2025-09-30, with the goal of preparing commercial uptake by the end of this period.
The project is led by a single German SME, SpiNNcloud Systems GmbH. With a 100% industry ratio and no university partners in the consortium, the project is heavily focused on commercial translation and productization rather than basic research.
Contact SpiNNcloud Systems GmbH in Germany for licensing and early access to SpiNNode boards.
Contact us to explore integration of neuromorphic hardware into your AI edge strategy.
