SubsidieMeestersReducing Carbon Footprint for Generative AI
The project aims to reduce the energy consumption and carbon footprint of generative AI systems by implementing more efficient matrix multiplication algorithms, potentially saving 40-50% energy while maintaining performance.
Projectdetails
Executive Summary
Introduction
Generative AI systems, such as chatGPT, recently passed the Turing test, forever transforming human-machine interaction. These systems provide giant productivity leaps across many sectors. However, their energy requirements increase nine-fold annually and their abundance grows at an exponential rate. The resulting carbon footprint becomes significant.
Industry Commitment
IT giants such as Google, Nvidia, Microsoft, and Amazon, as well as many mid-sized companies, have committed to reducing their carbon footprint. The EU is strengthening regulation for emission reductions. But the new generative AI trend jeopardizes emission reduction commitments.
Energy Consumption Challenges
Most power consumption of generative AI is spent on matrix multiplication. Our novel solutions reduce energy consumption and carbon footprint by replacing current matrix multiplication algorithms with more efficient ones. These can be implemented on existing hardware and software stacks.
Potential Savings
- Potential energy saving predicted at about 40-50%
- Maintains performance and accuracy
Research and Development
The novel developments of Prof. Oded Schwartz and his strong team are based on years of research and are protected by patents. The funds are requested to pursue business opportunities.
Financiële details & Tijdlijn
Financiële details
| Subsidiebedrag | € 150.000 |
| Totale projectbegroting | € 150.000 |
Tijdlijn
| Startdatum | 1-10-2023 |
| Einddatum | 31-3-2025 |
| Subsidiejaar | 2023 |
Partners & Locaties
Projectpartners
- THE HEBREW UNIVERSITY OF JERUSALEMpenvoerder
Land(en)
Vergelijkbare projecten binnen European Research Council
| Project | Regeling | Bedrag | Jaar | Actie |
|---|---|---|---|---|
Fast Matrix Multiplication for AIDeveloping patented methods for faster and energy-efficient matrix multiplication in software and hardware to enhance AI applications and capitalize on business opportunities. | ERC Proof of... | € 150.000 | 2023 | Details |
Outplaying the hardware lottery for embedded AIThe BINGO project aims to revolutionize embedded AI by enabling rapid customization of heterogeneous compute platforms using prefabricated chiplets, achieving 100x efficiency gains in days. | ERC Consolid... | € 1.995.750 | 2023 | Details |
Educational Inequalities and Generative AI: A Focus on Language DiversityChatEQUITY explores the impact of generative AI on educational inequalities by developing a new assessment tool and analyzing access disparities among students and teachers in diverse linguistic contexts. | ERC Consolid... | € 2.195.500 | 2025 | Details |
Growing Machines Capable of Rapid Learning in Unknown EnvironmentsGROW-AI aims to develop machines with general intelligence through genomic bottleneck algorithms and optimized learning environments, enhancing their autonomy and task-solving capabilities. | ERC Consolid... | € 1.994.225 | 2023 | Details |
Thermodynamic-inspired computing with oscillatory neural networksTHERMODON aims to revolutionize energy-efficient computing by integrating thermodynamics with neuromorphic architectures for self-organizing, adaptive AI systems. | ERC Consolid... | € 2.000.000 | 2024 | Details |
Fast Matrix Multiplication for AI
Developing patented methods for faster and energy-efficient matrix multiplication in software and hardware to enhance AI applications and capitalize on business opportunities.
Outplaying the hardware lottery for embedded AI
The BINGO project aims to revolutionize embedded AI by enabling rapid customization of heterogeneous compute platforms using prefabricated chiplets, achieving 100x efficiency gains in days.
Educational Inequalities and Generative AI: A Focus on Language Diversity
ChatEQUITY explores the impact of generative AI on educational inequalities by developing a new assessment tool and analyzing access disparities among students and teachers in diverse linguistic contexts.
Growing Machines Capable of Rapid Learning in Unknown Environments
GROW-AI aims to develop machines with general intelligence through genomic bottleneck algorithms and optimized learning environments, enhancing their autonomy and task-solving capabilities.
Thermodynamic-inspired computing with oscillatory neural networks
THERMODON aims to revolutionize energy-efficient computing by integrating thermodynamics with neuromorphic architectures for self-organizing, adaptive AI systems.
Vergelijkbare projecten uit andere regelingen
| Project | Regeling | Bedrag | Jaar | Actie |
|---|---|---|---|---|
First automated risk management platform to enable safety, fairness, explainability, and continuous monitoring of generative AI systemsQuantPis offers a comprehensive, model-agnostic platform for generative AI risk management, enabling companies to assess and mitigate risks while ensuring compliance with over 100 standards. | EIC Accelerator | € 2.498.475 | 2024 | Details |
Digital Business Architecture AssistantHet project onderzoekt de toepasbaarheid van generatieve AI voor het automatisch genereren van Business Architecture en Business Processes uit natuurlijke taal, om informatiesystemen efficiënter te ontwikkelen. | Mkb-innovati... | € 20.000 | 2023 | Details |
CO2-reductie in de glastuinbouw door intelligente algoritmenHet project ontwikkelt AI-gestuurde algoritmen voor het optimaliseren van het binnenklimaat in kassen, met als doel een CO2-reductie van 25% en optimale groeicondities. | Demonstratie... | € 341.050 | Onbekend | Details |
Plug & Play Artificial Energy IntelligenceHet project ontwikkelt een plug & play energiebesparingssoftware met kunstmatige intelligentie voor de industrie, gericht op CO2-reductie en kostenbesparing, getest in twee pilotomgevingen. | Demonstratie... | € 120.420 | Onbekend | Details |
n-ary spintronics-based edge computing co-processor for artificial intelligenceMultiSpin.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. | EIC Pathfinder | € 3.143.276 | 2024 | Details |
First automated risk management platform to enable safety, fairness, explainability, and continuous monitoring of generative AI systems
QuantPis offers a comprehensive, model-agnostic platform for generative AI risk management, enabling companies to assess and mitigate risks while ensuring compliance with over 100 standards.
Digital Business Architecture Assistant
Het project onderzoekt de toepasbaarheid van generatieve AI voor het automatisch genereren van Business Architecture en Business Processes uit natuurlijke taal, om informatiesystemen efficiënter te ontwikkelen.
CO2-reductie in de glastuinbouw door intelligente algoritmen
Het project ontwikkelt AI-gestuurde algoritmen voor het optimaliseren van het binnenklimaat in kassen, met als doel een CO2-reductie van 25% en optimale groeicondities.
Plug & Play Artificial Energy Intelligence
Het project ontwikkelt een plug & play energiebesparingssoftware met kunstmatige intelligentie voor de industrie, gericht op CO2-reductie en kostenbesparing, getest in twee pilotomgevingen.
n-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.