Skip to main content
Go to the home page of the European Commission (opens in new window)
English en
CORDIS - EU research results
CORDIS

Inferencing, Fast and Slow with Ultra-scaled Phase-Change Devices

Project description

Solutions for overcoming deep-learning inference challenges

The unsustainable energy demands of data movement are stalling AI’s progress, hindering its potential to contribute to economic and environmental advances. Conductance-based non-volatile memories can address this challenge by keeping synaptic weights stationary and enabling in-memory computation. However, existing devices remain far from their theoretical and functional limits. The ERC-funded INFUSED project bridges this gap by exploiting novel, ultra-scaled phase-change memory devices that efficiently encode both long-term and short-term adaptive weights, paving the way for a new generation of energy-efficient, biologically inspired AI hardware.

Objective

A major challenge for deep learning inference is the high energy demand required to retrieve large amounts of synaptic weight data from memory. One promising approach to address this is the use of conductance-based devices, such as non-volatile phase-change memory, to develop chips with stationary synaptic weights. However, two key obstacles remain: enhancing the computational capabilities and increasing the energy efficiency of these devices. INFUSED tackles both issues through groundbreaking device innovation. By utilizing the physics of ultra-scaled materials, it pushes energy efficiency closer to its theoretical limits. Moreover, it introduces dual neurally-plausible temporal dynamics, combining fast adaptive responses with slow, gradual conductance changes. This reimagines traditional neural network elements like the perceptron for more energy-efficient AI inference.

INFUSED specifically aims to:

☞ Develop a breakthrough device design method to encode slow weights efficiently by reducing electrical contacts and active volumes to unprecedented scales—below 10 nm², and using 1-nm-thick van der Waals films. These weights are programmable with energy in the tens of femtojoules, offering up to a 100x improvement over current industry devices.

☞ Design a reconfigurable volatile memory to encode fast weights within the slow weights. This innovative method uses field-effect properties of the ultra-scaled volumes. These weights exhibit non-linear temporal dynamics, consuming energy in the picoWatt range, with timescales spanning microseconds to hours, adaptable to AI tasks.

☞ Demonstrate a mixed hardware-software implementation of biologically inspired algorithms leveraging fast and slow dynamics. New neural architectures are developed and benchmarked against mainstream neural networks, as well as against hardware substrates like parallel computing architectures (e.g. GPUs), in complex vision tasks.

Fields of science (EuroSciVoc)

This project has not yet been classified with EuroSciVoc.
Be the first one to suggest relevant scientific fields and help us improve our classification service

Keywords

Programme(s)

Topic(s)

Funding Scheme

HORIZON-ERC - HORIZON ERC Grants

See all projects funded under this funding scheme

Call for proposal

(opens in new window) ERC-2025-STG

See all projects funded under this call

Host institution

IBM RESEARCH GMBH
Net EU contribution
€ 1 499 800,00
Address
SAEUMERSTRASSE 4
8803 RUESCHLIKON
Switzerland
Region
Schweiz/Suisse/Svizzera Nordwestschweiz Aargau
Activity type
Private for-profit entities (excluding Higher or Secondary Education Establishments)
Links
Total cost
€ 1 499 880,00

Beneficiaries (1)