As per the previews of the past few days, today NVIDIA announced a deal for acquire Hugging Face for $12.93 billionin an operation intended to further strengthen the company’s presence in the open artificial intelligence sector. The stated objective is increase the platform’s capacity, enhance its infrastructure and extend access to AI tools to developers and institutions around the world.
Hugging Face has become over the past decade one of the main points of reference for the community that develops and uses open artificial intelligence models. According to the data provided, the platform has over 18 million developers, researchers and creators, who share more than 3 million models, 500,000 datasets and 1 million applications. Furthermore, over 200,000 companies use Hugging Face to identify, evaluate, customize and deploy AI systems.
One of the aspects highlighted in the operation concerns the desire to keep the platform ecosystem open. Hugging Face will continue to allow developers to independently choose the models, frameworks, cloud services, inference providers and hardware platforms to use. In particular, the use of NVIDIA infrastructure will not be a requirement to develop or deploy applications through Hugging Face.
The platform It will also continue to support open source and open weight models from across the ecosystemregardless of their developer. Support for multi-cloud and multi-accelerator development and deployment will also remain, allowing designers to use the hardware and infrastructure best suited to their needs.
The question of open weight models is indicated as one of the central elements of the strategy. The announcement recalls a recent open letter, signed by industry representatives, on the importance of publicly accessible model weights for the economics of artificial intelligence. The principle advocated is that the availability of open models can broaden access to advanced AI capabilities and help distribute technological development across different companies, institutions and communities.
Open models allow startups, businesses, universities and public organizations to use advanced capabilities without having to train each system from scratch. They can also allow different models to be associated with specific activities, with potential applications ranging from cybersecurity to technological sovereignty, through industry, healthcare, agriculture, education and small businesses.
“AI advances faster when people can build together“, reads the announcement.
NVIDIA claims to already have a relevant role in the open Hugging Face ecosystem. The company claims to be the leading contributor of open data and models to the platform, with over 500 models and more than 250 datasets published. The company also publicly develops templates, libraries and tools that can be used, modified and extended by developers.
The acquisition should therefore make NVIDIA’s infrastructure, engineering resources and global presence available to Hugging Face to intervene on various aspects of the platform. The areas indicated include: reliability, security, model evaluation, inference, and deploymentwith the aim of supporting the growth of the ecosystem without changing the open character that contributed to the diffusion of Hugging Face.
“I’m honored that Clem came to me while considering the next chapter of Hugging Face and ha deemed NVIDIA a good home for the company, its community, and the future of open models” said Jense Huang, CEO of NVIDIA. “We share this vision, and the Hugging Face team will now bring their passion and expertise to a much bigger stage, while maintaining the same iconic brand 🤗.”
The operation is therefore part of a strategy that aims to combine NVIDIA’s infrastructure and computing capabilities with one of the main platforms dedicated to the sharing and development of AI models. “Together we will make AI more open, more capable, and more accessible to people and institutions around the world,” concludes Huang.
