Connect with us

International Circuit

Swisscom cooperates with NVIDIA to develop trusted generative AI solutions

Swisscom announced at AI House Davos a cooperation with NVIDIA to build generative AI (GenAI) full-stack supercomputers in Switzerland that will be based on NVIDIA accelerated computing and the NVIDIA AI Enterprise software platform. This includes the NVIDIA(opens in new tab) NeMo framework for building, customizing and deploying generative AI applications. Through the cooperation, Swisscom aims to develop standard, tailor-made and new AI use cases in collaboration with customers and for internal use.

As part of the cooperation, Swisscom plans to obtain the latest technologies from NVIDIA. Furthermore, Swisscom aims to establish a Trusted AI Factory, tapping the global ecosystem of NVIDIA, for Swiss customers.

Christoph Aeschlimann, Swisscom CEO says: “Working with NVIDIA enables us to act as a bridge-builder between the possibilities of technology and the needs of customers. As a first step, we are building a high-performance infrastructure upon which we will create trustworthy services. We have been in close contact with many customers to consult them on their specific needs.”

Swisscom Group plans to invest an amount of up to 100 million Swiss Francs related to AI Infrastructure and service development in the coming years.

“Telcos are increasingly seen as leaders in building, hosting and deploying sovereign AI. Having a trusted, high-performance infrastructure is becoming critical to every nation’s technological advancement,” said Keith Strier, Vice President of Worldwide AI Initiatives at NVIDIA. “Building on NVIDIA’s full-stack accelerated computing platform for streamlining the development and deployment of production-grade AI applications Swisscom can enable leading, real-world AI solutions to Swiss companies, driving the transformation of the country’s economy.”

AI for Swiss Customers
When using publicly available generative AI services, the methods of data handling can often be intransparent and difficult for companies to comprehend. Christoph Aeschlimann from Swisscom says: “Everyone talks about generative AI, but the gap between expectations and reality is tremendous. A customer must be sure what happens with its data, where it happens and how it happens. Switzerland, with its many multinational organizations, needs trusted and unique sovereign building blocks for AI. That is where we step in. We are developing it in a reliable and secure manner, with a distinct Swiss flavor, hand in hand with NVIDIA.”

With this collaboration, Swisscom gains access to NVIDIA industry experts who have direct experience in implementing AI to various industries. Swisscom also now stands as a strategic reseller of NVIDIA technology, including the NVIDIA DGX platform and NVIDIA GPUs. With NVIDIA as its technology partner, Swisscom is poised to become a comprehensive one-stop shop for all AI customer needs in Switzerland.

Swisscom has been using AI technologies within its own operations for several years in many fields such as customer service, network operations and other services. Swisscom started early with optimizing internal operations, e.g. ticket classification and automatic triage. It generated its own Swiss-German language model to provide better customer service, e.g. on its TV platform with voice search. Swisscom is using augmented reality in TV studios for football kick-off detection and machine learning to optimize and detect anomalies in its networks.

As part of the group AI strategy, Fastweb, a Swisscom Italian subsidiary, announced last December the purchase of 31 NVIDIA DGX H100(opens in new tab) systems, the first large-scale NVIDIA DGX AI-powered supercomputer in Italy for the development of a national AI system. Fastweb will make available the highly performant NVIDIA DGX H100 cluster in the cloud as an infrastructure-as-a-service and will leverage it to develop and deploy the first large language model natively trained in the Italian language.

CT Bureau

Click to comment

You must be logged in to post a comment Login

Leave a Reply

Copyright © 2024 Communications Today

error: Content is protected !!