Nvidia Collaborates with Hugging Face for Innovative Cloud-Based AI Training Solutions

Nvidia is joining forces with AI startup Hugging Face to enhance access to AI computing power. Coinciding with the annual SIGGRAPH conference this week, Nvidia announced its support for Hugging Face's new service, Training Cluster as a Service, designed to streamline the creation of custom generative AI models for businesses.

Scheduled for launch in the coming months, Training Cluster as a Service will leverage Nvidia's DGX Cloud, a comprehensive AI "supercomputer" hosted in the cloud. This robust platform provides access to an instance equipped with eight Nvidia H100 or A100 GPUs and 640GB of GPU memory. It also includes Nvidia’s AI Enterprise software, which helps develop AI applications and large language models, alongside consultations with Nvidia experts.

While companies can directly subscribe to DGX Cloud, starting at $36,999 per month per instance, Training Cluster as a Service seamlessly integrates this infrastructure with Hugging Face’s extensive library of over 250,000 models and more than 50,000 datasets—an invaluable resource for AI initiatives.

“People worldwide are forging new connections and making discoveries with generative AI tools, and we’re still in the early stages of this technological evolution,” said Hugging Face co-founder and CEO Clément Delangue. “Our partnership with Nvidia will harness their cutting-edge AI supercomputing technology to empower businesses to shape their AI journey. We aim to support the open source community by providing the software and resources they need to contribute to future advancements.”

This collaboration comes as Hugging Face reportedly seeks to secure additional funding at a valuation of $4 billion. Founded in 2014 by Delangue, Julien Chaumond, and Thomas Wolf, Hugging Face has rapidly evolved from a consumer application to a comprehensive resource for AI model development. According to Delangue, the platform currently supports over 15,000 organizations.

The partnership is a strategic move for Nvidia, which has significantly expanded its cloud services for training, experimenting with, and deploying AI models in response to rising demand. Earlier this year, the company introduced AI Foundations, a suite of tools enabling developers to create custom generative AI models tailored for specific applications.

AI is forecasted by tech market research firm Tractica to make up to 50% of total public cloud services revenue by 2025. The demand for AI cloud training infrastructure is so intense that it has led to hardware shortages, prompting cloud providers like Microsoft to temper growth expectations.