Fastweb Selects NVIDIA DGX H100 systems to Pioneer Native Italian AI Language Model

Fastweb, the Italian telecoms operator, has unveiled a major initiative aimed at revolutionizing AI development in Italy.
Generative AI chips NVIDIAThe company announced its strategic plan to construct an Artificial Intelligence Large-Language Model (LLM) specifically trained in native Italian. This ambitious project aims to empower corporate and public administration clients in crafting advanced AI tools.

The AI project will see a team of generative AI experts leveraging deep learning systems to construct an LLM grounded in extensive datasets of the Italian language. This innovative approach by Fastweb aims to cater to the unique linguistic nuances and intricacies of Italian, fostering a more robust foundation for AI tool development.

Fastweb will purchase 31 NVIDIA DGX H100 systems. These systems, incorporating 248 NVIDIA H100 Tensor Core GPUs, are slated to power one of Fastweb’s forthcoming data centers in Lombardy, operating on renewable energy sources, by the first half of 2024.

The implementation of the NVIDIA DGX SuperPOD reference architecture within these systems positions Fastweb at the forefront of artificial intelligence capabilities. These resources will be harnessed as Infrastructure as a Service (IaaS) in the cloud, enabling customers to develop AI and generative AI applications.

This bespoke LLM will serve as a dedicated platform, fostering the creation of a diverse spectrum of generative AI applications by Fastweb and third-party developers. Fastweb’s CEO, Walter Renna, highlighted the significance of this venture, expressing how it positions Fastweb as a catalyst for Italy’s digital transformation by offering a robust, transparent, and sustainable framework for AI application development.

Ronnie Vasishta, SVP of Telecom at NVIDIA, commended Fastweb’s forward-looking approach, emphasizing the pivotal role telcos will play in the cloud computing era. Vasishta views Fastweb’s investment as a pivotal step in providing powerful, efficient, and localized computing resources to AI innovators in Italy.

Fastweb’s statement emphasized the significance of this new LLM, positioning it as a nation-specific platform facilitating the creation of an extensive array of generative AI applications. By offering this advanced infrastructure and specialized AI platform, Fastweb aims to accelerate innovation both within its own operations and among external collaborators.

Fastweb’s ambitious move aligns with the evolving landscape of AI technology and showcases its commitment to fostering localized innovation and technological advancements in Italy. This initiative holds the promise of unlocking new potentials in AI application development, fostering a new era of technological innovation driven by the intricacies of the Italian language and market dynamics.