SubsidieMeestersn-ary spintronics-based edge computing co-processor for artificial intelligence
MultiSpin.AI aims to revolutionize edge computing by developing a neuromorphic AI co-processor that enhances energy efficiency and processing speed, enabling transformative applications while reducing CO2 emissions.
Projectdetails
Introduction
The rise of technologies such as the Internet of Things (IoT), autonomous vehicles, smart cameras, etc. is generating lots of big data. The volume of data in 2022 was 97ZB and is doubling every 2-3 years. This is leading to unprecedented growth in energy consumption and costs needed for data processing.
Challenges in Data Processing
Sending raw data for remote processing on centralized nodes is limited in terms of speed and bandwidth. Even next-gen tech like 5G or 6G will be insufficient to cope with this growth. Processing data at the Edge, where it's generated, requires increasing power efficiency by several orders of magnitude.
Limitations of Current Technology
However, the use of general-purpose digital processors based on von Neumann architecture is limited, with optimization possibilities nearing natural limits.
The Need for Neuromorphic Hardware
A new class of chips, neuromorphic hardware, is needed to execute AI algorithms like Deep Learning at high speed, low energy consumption, endurance, and scalability.
MultiSpin.AI's Vision
MultiSpin.AI's vision is to improve neuromorphic computing by:
- Increasing energy efficiency and processing speed by at least three orders of magnitude over digital computing.
- Achieving more than 10x improvement compared to the most advanced neuromorphic devices.
- Reaching an unparalleled 2,000 Tera operations per second per watt (TOPS/W).
Development of AI Co-Processor
To achieve this, MultiSpin.AI will develop an AI co-processor based on a crossbar of multi-level magnetic tunnel junctions (M2TJ) cells/n-ary state cells.
Advantages of Multi-Level M2TJs
The use of multi-level M2TJs offers several benefits:
- Reduces the number of cells.
- Simplifies circuitry.
- Reduces the number of digital-to-analog conversions (DAC) at the input of the crossbar.
- Reduces analog-to-digital conversions at the crossbar output.
The combined effect is realizing much higher energy efficiency and faster AI inference at the Edge.
Impact of the Breakthrough
This breakthrough will help provide a significant impact by enabling transformative applications like:
- Autonomous vehicles
- Robots
- Medical devices
It will also help strengthen strategic autonomy for the EU chips industry and reduce CO2 emissions from AI inference.
Financiële details & Tijdlijn
Financiële details
| Subsidiebedrag | € 3.143.276 |
| Totale projectbegroting | € 3.143.276 |
Tijdlijn
| Startdatum | 1-2-2024 |
| Einddatum | 31-1-2027 |
| Subsidiejaar | 2024 |
Partners & Locaties
Projectpartners
- BAR ILAN UNIVERSITYpenvoerder
- INESC MICROSISTEMAS E NANOTECNOLOGIAS - INSTITUTO DE ENGENHARIA DE SISTEMAS E COMPUTADORES PARA OS MICROSISTEMAS E AS NANOTECNOLOGIAS
- UNIVERSITE CATHOLIQUE DE LOUVAIN
- SPINEDGE LTD
- INTERACTIVE FULLY ELECTRICAL VEHICLES SRL
- VRIJE UNIVERSITEIT BRUSSEL
- AMIRES SRO
Land(en)
Vergelijkbare projecten binnen EIC Pathfinder
| Project | Regeling | Bedrag | Jaar | Actie |
|---|---|---|---|---|
Hybrid electronic-photonic architectures for brain-inspired computingHYBRAIN aims to develop a brain-inspired hybrid architecture combining integrated photonics and unconventional electronics for ultrafast, energy-efficient edge AI inference. | EIC Pathfinder | € 1.672.528 | 2022 | Details |
SPIKING PHOTONIC-ELECTRONIC IC FOR QUICK AND EFFICIENT PROCESSINGSPIKEPro aims to develop an integrated neuromorphic chip combining electrical and photonic neurons to create efficient, high-speed spiking neural networks for diverse applications. | EIC Pathfinder | € 1.973.038 | 2024 | Details |
Green SELf-Powered NEuromorphic Processing EnGines with Integrated VisuAl and FuNCtional SEnsingELEGANCE aims to develop eco-friendly, light-operated processing technology for energy-efficient IoT applications, utilizing sustainable materials to minimize electronic waste and environmental impact. | EIC Pathfinder | € 3.100.934 | 2024 | Details |
RECONFIGURABLE SUPERCONDUTING AND PHOTONIC TECHNOLOGIES OF THE FUTURERESPITE aims to develop a compact, scalable neuromorphic computing platform integrating vision and cognition on a single chip using superconducting technologies for ultra-low power and high performance. | EIC Pathfinder | € 2.455.823 | 2023 | Details |
Digital optical computing platform for neural networksDOLORES aims to develop a digital optical neural network processor to overcome current optical computing limitations, revolutionizing AI and deep learning applications across various sectors. | EIC Pathfinder | € 3.015.883 | 2024 | Details |
Hybrid electronic-photonic architectures for brain-inspired computing
HYBRAIN aims to develop a brain-inspired hybrid architecture combining integrated photonics and unconventional electronics for ultrafast, energy-efficient edge AI inference.
SPIKING PHOTONIC-ELECTRONIC IC FOR QUICK AND EFFICIENT PROCESSING
SPIKEPro aims to develop an integrated neuromorphic chip combining electrical and photonic neurons to create efficient, high-speed spiking neural networks for diverse applications.
Green SELf-Powered NEuromorphic Processing EnGines with Integrated VisuAl and FuNCtional SEnsing
ELEGANCE aims to develop eco-friendly, light-operated processing technology for energy-efficient IoT applications, utilizing sustainable materials to minimize electronic waste and environmental impact.
RECONFIGURABLE SUPERCONDUTING AND PHOTONIC TECHNOLOGIES OF THE FUTURE
RESPITE aims to develop a compact, scalable neuromorphic computing platform integrating vision and cognition on a single chip using superconducting technologies for ultra-low power and high performance.
Digital optical computing platform for neural networks
DOLORES aims to develop a digital optical neural network processor to overcome current optical computing limitations, revolutionizing AI and deep learning applications across various sectors.
Vergelijkbare projecten uit andere regelingen
| Project | Regeling | Bedrag | Jaar | Actie |
|---|---|---|---|---|
SpiNNaker on the EdgeSpiNNcloud Systems aims to develop real-time, energy-efficient AI applications by transitioning cutting-edge neuromorphic computing technology from cloud to edge, enhancing performance and commercialization. | EIC Transition | € 2.499.998 | 2023 | Details |
Memristive Neurons and Synapses for Neuromorphic Edge ComputingMEMRINESS aims to develop compact, power-efficient Spiking Neural Networks using memristive technology for enhanced collaborative learning on edge systems. | ERC Starting... | € 1.499.488 | 2022 | Details |
Artificial Intelligence–Driven Materials Design for Spintronic ApplicationsThis project aims to develop AI tools to optimize Van der Waals heterostructures for energy-efficient spin-orbit torque memories, enhancing speed and storage while reducing power consumption. | ERC Starting... | € 1.078.750 | 2023 | Details |
Hybrid Spintronic Synapses for Neuromorphic ComputingSpin-Ion Technologies aims to develop neuromorphic chips using ion beam-engineered magnetic materials, bridging computational neuroscience and deep learning for efficient embedded systems. | EIC Transition | € 2.499.998 | 2023 | Details |
Scalable Unified Processor Enhancing Revolutionary Computing, Harnessing Integrated Performance for Edge AI, Autonomous Driving, Generative AI, and Decentralized AIoT ApplicationsDeveloping the Tyr chip to enable real-time, efficient processing for Level 4/5 autonomous driving, addressing current data and processing limitations in the industry. | EIC Accelerator | € 2.499.999 | 2023 | Details |
SpiNNaker on the Edge
SpiNNcloud Systems aims to develop real-time, energy-efficient AI applications by transitioning cutting-edge neuromorphic computing technology from cloud to edge, enhancing performance and commercialization.
Memristive Neurons and Synapses for Neuromorphic Edge Computing
MEMRINESS aims to develop compact, power-efficient Spiking Neural Networks using memristive technology for enhanced collaborative learning on edge systems.
Artificial Intelligence–Driven Materials Design for Spintronic Applications
This project aims to develop AI tools to optimize Van der Waals heterostructures for energy-efficient spin-orbit torque memories, enhancing speed and storage while reducing power consumption.
Hybrid Spintronic Synapses for Neuromorphic Computing
Spin-Ion Technologies aims to develop neuromorphic chips using ion beam-engineered magnetic materials, bridging computational neuroscience and deep learning for efficient embedded systems.
Scalable Unified Processor Enhancing Revolutionary Computing, Harnessing Integrated Performance for Edge AI, Autonomous Driving, Generative AI, and Decentralized AIoT Applications
Developing the Tyr chip to enable real-time, efficient processing for Level 4/5 autonomous driving, addressing current data and processing limitations in the industry.