Design and Integration of a Multi-Precision Systolic GEMM Accelerator in RISC-V Vector Clusters via AME Extensions for Edge AI Applications

Scientific-Disciplinary Group

09/IINF-01 - Electronics

Description

This Incarico di Ricerca position, aligned with the ARCHYTAS project, focuses on the design and evaluation of matrix-acceleration solutions for AI workloads. The research addresses: (1) the design of a multi-precision floating-point systolic GEMM accelerator for AI inference and fine-tuning on edge SoCs; (2) its integration as a co-processor within a RISC-V vector cluster through RISC-V AME extensions, with attention to architectural interfacing and software accessibility; and (3) functional validation, plus performance, power, and area characterization and benchmarking on representative AI kernels. The goal is to advance the state of the art in open, efficient, and portable AI acceleration for multi-tile heterogeneous compute systems, contributing to European technological sovereignty and to design methodologies consistent with the ARCHYTAS project roadmap.

Job posting website

https://bandi.unibo.it/ricerca/incarichi-di-ricerca

Funding body

ALMA MATER STUDIORUM - UNIVERSITA' DI BOLOGNA - - DIPARTIMENTO DI INGEGNERIA DELL'ENERGIA ELETTRICA E DELL'INFORMAZIONE "GUGLIELMO MARCONI"

How to apply

Other

Selection process