Generative AI continues to make headlines, and ChatGPT has thrust debates about the power and potential of AI back into popular culture, but corporate enterprises do not need popular chatbots as much as they need practical AI tools to help them leverage increasing amounts of data into valuable business tools and applications.
With a new package of services announced at its Spring GTC conference this week, Nvidia is looking to answer that need. The company unveiled AI Foundations, a set of cloud-based “model-making” services that enable businesses to put AI to work on their own terms and for their own purposes. The platform can help them build, refine, and operate custom large language models and Generative AI models that are trained with their own proprietary data–including language, image, video, and 3D content–and created for their unique domain-specific tasks, according to Manuvir Das, vice president of enterprise computing at Nvidia.
The unveiling was part of a news-packed day at GTC, and one of many announcements aimed at helping enterprises take advantage of what Nvidia CEO Jensen Huang has described as “the iPhone moment for AI.”
Das made clear that AI Foundations is not intended to be yet another Generative AI chatbot service in the mold of ChatGPT. “Generative AI has been in the news so much in the last few months,” he said. “This is not a new Generative AI offering from Nvidia. Rather, this is a platform, a foundry that allows companies to create their own models that they can then use in their own Generative AI services.”
Nvidia also is offering pre-trained models “and the idea there is to give a company a head start so they don't have to train models from scratch,” Das said, but the focus of AI Foundations is “customization,” he added. He emphasized, under questioning from reporters during a pre-GTC briefing, that Nvidia is not intending to compete with OpenAI, Google, or others, and that it is trying to open up the ability to train models for Generative AI from the likes of these giants with deep pockets and armies of engineers to individual enterprises with their own evolving use cases and customer groups that can benefits from them.
The first users of AI Foundations include big name firms with huge data repositories – Adobe, Getty Images, Morningstar, Quantiphi and Shutterstock.
Nvidia’s NeMo language service and its Picasso image, video and 3D service are components of the AI Foundations platform that enterprises can use to build proprietary, domain-specific, generative AI applications for intelligent chat and customer support, professional content creation, digital simulation and more. Nvidia also announced at GTC that it will support new models for the NVIDIA BioNeMo cloud service for biology.
Both NeMo and Picasso run on the new DGX Cloud, a cloud-based version of the Nvidia DGX supercomputing system for AI that Nvidia separately announced at GTC, and which will give enterprises browser-based API access to the models. Once models are ready for deployment, enterprises can run inference workloads at scale using the AI Foundations cloud services. Each cloud service includes six elements: pretrained models, frameworks for data processing, vector databases and personalization, optimized inference engines, APIs, and support from NVIDIA experts to help enterprises tune models for their custom use cases.
Regarding the Adobe relationship, Nvidia and Adobe said AI Foundations will be used to jointly develop models, some of which will be marketed through Adobe products like Photoshop, Premiere Pro, and Adobe After Effects, as well as through Nvidia Picasso. Picasso also will host some models from Adobe Firefly, a new family of creative generative AI models, the first of which Adobe separately announced beta, and which is focused on the generation of images and text effects designed to be safe for commercial use.
“Adobe and Nvidia have a long history of working closely together to advance the technology of creativity and marketing,” said Scott Belsky, Chief Strategy Officer and EVP, Design and Emerging Products, Adobe. “We’re thrilled to partner with them on ways that generative AI can give our customers more creative options, speed their work, and help scale content production.”
Meanwhile, Shutterstock acknowledged that it is working with the Picasso service to train custom 3D models to create generative 3D assets from simple text prompts. That will streamline creation time from hours to minutes, according to Shutterstock, which plans to introduce the models in the coming months on Shutterstock.com.
“Our generative 3D partnership with Nvidia will power the next generation of 3D contributor tools, greatly reducing the time it takes to create beautifully textured and structured 3D models,” said Shutterstock CEO Paul Hennessy. “This first of its kind partnership furthers our strategy of leveraging Shutterstock’s massive pool of metadata to bring new products, tools, and content to market. By combining our 3D content with Nvidia’s foundation models, and utilizing our respective marketing and distribution platforms, we can capitalize on an extraordinarily large market opportunity.”
Aside from the AI Foundations news, Das also discussed how more enterprises in more industries are starting to embrace AI tools. For example, Nvidia this week announced telecom giant AT&T, a longtime Nvidia partner, as a new user of the Nvidia AI Enterprise platform.
“Industries are embracing a new era in which chatbots, recommendation engines and accelerated libraries for data optimization help produce AI-driven innovations,” Das said. “Our work with AT&T will help the company better mine its data to drive new services and solutions for the AI-powered telco.”
He added, “I’m highlighting this relationship with AT&T because this is how we expect to work with the largest enterprise companies in the world going forward – by offering one platform, one architecture that they can standardize on and benefit from for different AI use cases.”
RELATED: Nvidia software release aims to spur enterprise AI expansion