Proof of Concept Prototype for Matching Applications on BELFORT Hardware

Informations projet

BELFORT MATCH

N° de convention de subvention: 101158183

DOI

10.3030/101158183

Projet clôturé

Date de signature de la CE 15 Février 2024

Date de début 1 Mars 2024

Date de fin 31 Août 2025

Financé au titre de

• European Research Council (ERC)

Coût total

Aucune donnée

Contribution de l’UE

€ 150 000,00

Coordonné par

KATHOLIEKE UNIVERSITEIT LEUVEN
Belgium

Periodic Reporting for period 1 - BELFORT MATCH (Proof of Concept Prototype for Matching Applications on BELFORT Hardware)

Période du rapport: 2024-03-01 au 2025-08-31

Résumé du contexte et des objectifs généraux du projet

The project fits into the context of designing and developing hardware to accelerate computations on encrypted data (COED). In this proof of concept we focus on the Fully Homomorphic Encryption (FHE) approach to COED. The ultimate goal is to make the operations on encrypted data efficient enough to perform them on the fly in a cloud setting. When a user formulates privacy-preserving encrypted queries to a server, it should be imperceptible to the user that the operations are occurring in the encrypted domain. Being able to answer a query on the fly without delay will make this goal happen.
With this POC application, we demonstrate our progress on the hardware acceleration of encrypted computation. The POC is used for two main efforts consisting of a technical part and an innovation support part. The goal of the technical part is to raise the Technology Readiness Level (TRL) of the BELFORT accelerator, integrate it into software libraries and develop an actual ‘match’ demonstrator. The innovation support part will focus on making the project business ready.

Travail effectué depuis le début du projet jusqu’à la fin de la période considérée dans le rapport et principaux résultats atteints jusqu’à présent

On the technical and scientific aspects, the proof of concept has one major main achievement. Within the ERC POC, a public demonstrator for encrypted string matching calculations, which we call Leuvenshtein, was developed. Edit distance computations are essential in applications across finance and genomics, such as DNA sequence alignment. The public demonstrator runs on the FPGA hardware platforms developed within the ERC Grant Belfort, achieving a 278x speedup. The demonstrator has high potential to attract business partners for use-cases across finance and genomics. It has been open-sourced in a public repository under an MIT license. The demonstrator has also been published at Usenix Security 2025, an A* conference on security and privacy.

Progrès au-delà de l’état des connaissances et impact potentiel prévu (y compris l’impact socio-économique et les conséquences sociétales plus larges du projet jusqu’à présent)

Through the participation at key industry events, the POC provided also market insights for computing on encrypted data. Based on the results of the ERC grant Belfort, the spin-off company BelfortLabs has been created. Besides the development of the demonstrator, the Proof of concept grant was used to understand the market and to prepare for investor seed funding. Moreover, the proof of concept provides for an entrance ticket for a follow-up proposal for an EIT Transition grant.