Skip to main content
Aller à la page d’accueil de la Commission européenne (s’ouvre dans une nouvelle fenêtre)
français fr
CORDIS - Résultats de la recherche de l’UE
CORDIS

Synchronous Optical Network Infrastructure for the Cloud

Objectif

Cloud data center networks (DCNs) are a key enabler of the digital world. The explosive rise of artificial intelligence (AI) is pushing them to unprecedented bandwidth, scale, energy, and cost limits. As electrical switching approaches fundamental scaling and power barriers, simply building ever larger packet-switched networks is becoming unsustainable. Meanwhile, a new wave of optical technologies—such as silicon photonics, co-packaged optics, Tbps optical I/O, and nanosecond-scale optical circuit switching—is rapidly transforming how data can move throughout the AI computing fabric.

Training and serving large AI models usually involve thousands of accelerators executing highly coordinated synchronous computations with repetitive and predictable communication patterns. This creates a unique opportunity: rather than forcing AI workloads onto networks designed for traditional unpredictable traffic, can we redesign the network around the structure of AI computation?

SONIC will pioneer optical network infrastructure for AI computing in the cloud. It will exploit dynamically reconfigurable, high-capacity optical circuits to tightly orchestrate computation and communication across large-scale AI clusters. SONIC will develop new mechanisms to integrate AI accelerators with emerging optical technologies, co-schedule optical circuits with AI collective communication, bridge circuit-switched optics with existing network protocols, and maintain correctness and performance under failures and optical impairments.

By transforming optics from passive connectivity into an active part of the computing system, SONIC aims to break through the bandwidth and energy barriers of electrical networks and establish a new foundation for efficient AI infrastructure. This project would enable increasingly powerful AI systems, fundamentally reshape the future computing landscape, and substantially reduce the DCN energy footprint.

Champ scientifique (EuroSciVoc)

Ce projet n'a pas encore été classé par EuroSciVoc.
Proposez les domaines scientifiques qui vous semblent les plus pertinents et aidez-nous à améliorer notre service de classification.

Mots‑clés

Programme(s)

Thème(s)

Régime de financement

HORIZON-ERC - HORIZON ERC Grants

Voir tous les projets financés dans le cadre de ce programme de financement

Appel à propositions

(s’ouvre dans une nouvelle fenêtre) ERC-2026-STG

Voir tous les projets financés au titre de cet appel

Institution d’accueil

MAX-PLANCK-GESELLSCHAFT ZUR FORDERUNG DER WISSENSCHAFTEN EV
Contribution nette de l'UE
€ 1 499 983,00
Coût total
€ 1 499 983,00

Bénéficiaires (1)