At the event organised by Do IT Now in Turin, Riccardo Di Maria, Head of HPC/AI and Engineering at AI4I, discussed the evolution of advanced computing infrastructures, the convergence of HPC and AI, and the growing strategic importance of specialised skills.
On 25 September 2026, the Italian Institute of Artificial Intelligence (AI4I) took part in HPC Roundtable 26, organised by Do IT Now in Turin, bringing together experts, companies and technology leaders to discuss the evolution of High Performance Computing in a technology landscape increasingly shaped by artificial intelligence.
Representing AI4I, Riccardo Di Maria, Head of HPC/AI and Engineering, contributed to the discussion on some of the main transformations affecting the sector: from hybrid architectures and the relationship between traditional HPC and AI Factories to changing business requirements and the dynamics of an increasingly concentrated global technology market.
HPC and AI: infrastructures are converging
As AI workloads become increasingly central to advanced computing, the boundaries between traditional HPC systems and AI infrastructure are becoming less distinct.
The discussion addressed how organisations can manage increasingly hybrid infrastructures, combining different computing environments and technologies, and whether the distinction between HPC and AI Factories remains useful as architectures, workloads and operational requirements increasingly converge.
For AI4I, this evolution is particularly relevant to the development of computing capabilities designed to support advanced research, engineering and industrial applications.
New requirements from industry
The transformation is also changing the way companies approach computing infrastructure.
The Roundtable examined the choices organisations face between proprietary HPC and AI capacity, public cloud resources and managed services, as well as the need to build infrastructure models capable of adapting to different workloads and business requirements.
The challenge is therefore not only to increase computing capacity, but to connect infrastructure choices to concrete research, engineering and industrial needs.
Infrastructure needs talent
One point emerged particularly strongly from the discussion: infrastructure alone is not enough without the people able to design, operate and continuously evolve it.
Global competition for highly specialised HPC and AI professionals makes the availability of technical expertise an increasingly strategic issue. Building advanced computing capacity therefore also means attracting, developing and retaining the skills required to manage complex infrastructures over time.
In this context, AI4I is working to bring together expertise, computing infrastructure and collaboration with research and industry, contributing to the development of the technological capabilities needed to address the next generation of AI and HPC challenges.
