The HiCREST Laboratory is proud to announce that both reference professors of our research group, Prof. Paolo Rech and Prof. Flavio Vella, have secured funding under the PRIN 2026 call issued by the Italian Ministry of University and Research (MUR).
These two major grants recognize the scientific value and high potential impact of our ongoing research lines in High-Performance Computing (HPC), AI reliability, and sustainable computing environments.
⚡ CRESCENDO: Carbon-Reducing, Energy-Efficient, Sustainable Computing ENvironments for Datacentre Operations
(PRIN 2026 HYBRID — Research Unit Lead: Prof. Flavio Vella)
With a total grant of €1.2 million, the CRESCENDO project addresses the growing environmental, water, and energy footprint driven by modern datacenter-scale HPC simulations and AI workloads.
The initiative adopts an interdisciplinary approach structured around three core pillars:
- Energy: Energy-efficient system software for compute, storage, and networking under HPC and AI workloads.
- Carbon: Carbon-aware datacenter infrastructure, including thermal control and advanced water cooling.
- Sustainability: Lifecycle assessment, embodied carbon, and territorial environmental impact.
These pillars will converge into a multi-layer digital twin connecting computing efficiency directly to datacenter infrastructure. All resulting tools, models, and datasets will be released as open-source resources in alignment with the European Green Deal.
Collaborating Institutions: University of Padova, University of Trento, Politecnico di Milano, Area Science Park Trieste.
🛡️ AEGIS: Ensuring AI Reliability and Longevity Through Hardware and Software Cross-Layer Hardening
(PRIN 2026 — Consortium and Research Unit Lead: Prof. Paolo Rech)
As artificial intelligence systems are rapidly adopted across industrial domains, the advanced hardware powering modern models becomes increasingly vulnerable to reliability faults. The AEGIS project addresses this challenge by engineering hardware- and software-hardened AI systems that remain resilient against underlying hardware failures during both inference and training phases.
By establishing cross-layer hardening mechanisms, AEGIS aims to enable safe and dependable AI deployment in safety-critical sectors—such as autonomous driving and space exploration—where hardware reliability risks currently limit broader AI adoption.
Collaborating Institutions: University of Trento (DII), Politecnico di Milano, Politecnico di Torino, Alma Mater Studiorum – Università di Bologna.
🎓 Upcoming Opportunities
In the coming months, both research initiatives will open new academic positions. Further announcements regarding recruitment and project progress will be published on our website and social channels.
