A Strategic
Infrastructure
Designed to combine high-performance computing capabilities, engineering expertise and specialised services, itsupports the entire project lifecycle, from prototype to industrial adoption.
Researchers, startups, SMEs and large enterprises can access an operational environment designed to develop, test, validate and scale AI solutions quickly, securely and in line with European standards. A strategic infrastructure connecting applied research, industrial innovation and technological autonomy.
Turning AI into
Industrial Impact
AI Foundry Peano is AI4I’s computational infrastructure dedicated to the development, experimentation and deployment of artificial intelligence applications at industrial scale.
Not a traditional research laboratory, but an operational platform where AI projects move through their entire lifecycle: from prototyping to validation, through to production. A pilot line designed to bridge applied research and industrial adoption.
Beyond the
experimental phase
The Foundry is designed to address a concrete need: moving AI beyond the experimental phase and into real integration within industrial processes and services. It does so by bringing together three elements that traditionally remain separate: high-performance computing power, specialist engineering expertise, and a governance framework ensuring security, reliability and regulatory compliance. Those who access the Foundry gain not just computational resources, but a managed infrastructure designed to support projects from inception through to production.
148 next-generation GPUs
80 NVIDIA B200 68 NVIDIA H200 advanced AI and HPC workloads
The computational infrastructure is built on 80 B200 and 68 H200 NVIDIA GPUs, designed to support training, simulation, inference, validation and large-scale dataset management for advanced AI and HPC workloads.
An outstanding computing resource for researchers and organizations
Flexible access models and ready-to-use computational environments
68
NVIDIA H200
For generative AI and high-performance computing (HPC) workloads
80
NVIDIA B200
Next-generation datacenter GPU based on the Blackwell architecture
One platform
for AI and HPC
AI Foundry Peano integrates AI and HPC capabilities, a unified cloud platform, and advanced orchestration, scheduling and virtualization environments, including Slurm-based resource management systems, OpenStack, Virtual Machines, Kubernetes and interactive developmentenvironments. Dedicated sandboxing and testing environments support experimentation, optimization and debugging activities across the full development lifecycle.
A broad range
of users
Designed to serve a broad range of users, from academic researchers to companies and industrial partners, the Foundry provides flexible access models and ready-to-use computational environments that reduce setup complexity and accelerate time-to-result.
A significant share of the effective computing capacity is allocated to research and development projects carried out by AI4I’s Research, Development and Engineering Laboratories, spanning internal research, joint labs, commissioned research and consortium activities. The Foundry also serves as a key operational resource for the SUK, another core pillar of AI4I, with dedicated access conditions and usage incentives in place.
The remaining computational resources are available to external users and organizations, including private entities, SMEs and industrial partners.
Designed for Strategic Autonomy
Within AI4I’s strategic framework, the Foundry is positioned as a pilot infrastructure supporting future scale operations on larger systems, including AI Factory IT4LIA and AI National Factories. It is one of the Institute’score infrastructural assets, connecting applied research, engineering capability and industrial adoption.
Artificial intelligence has become strategic infrastructure, on a par with energy or communications networks. A national platform means developing and governing AI according to European standards, in directresponse to the needs of Italian industry, and reducing dependence on external infrastructures and paradigms.
For Italy, Foundry Peano is a foundation for autonomy, competitiveness and industrial innovation.
A category of its own
Hyperscalers offer computational power at global scale, but within proprietary ecosystems governed by the provider’s rules, pricing and constraints. Large national supercomputers are optimised for pure scientificresearch, with access models that do not always align with the needs and timelines of industry.
Foundry Peano occupies a distinct space: a managed infrastructure designed to be accessible to research laboratories and companies without requiring advanced infrastructure expertise. Ready-to-use environments, integrated technical and engineering support, security and regulatory compliance built in from the outset.
The competitive advantage lies not only in raw computing power, but in the capacity to iterate rapidly: develop, test, validate and deploy in short cycles, with full process control.
