Nvidia Launches Enterprise Generative AI Cloud Services for Synthetic Media Engines

Nvidia introduced a collection of generative AI enterprise cloud services this week at their annual GTC event. The Nvidia AI Foundations software portfolio is designed to allow companies to create custom, domain-specific, generative AI applications around synthetic media and interactive AI personalities. Neither service is widely available yet, and companies have to apply for access.

NeMo Picasso

The cloud services aim to bring generative AI models trained on proprietary data and built for specific purposes to businesses that don’t want to build their own from scratch. The Nvidia AI Foundations services are built on the Nvidia NeMo model for language and Nvidia Picasso for synthetic media, including still images, videos, and three-dimensional models. The package includes pre-trained models and software support data processing and APIs. Once the custom models are already, they can be embedded in software as an API and run at scale. NeMo offers developers the option to include custom data for generating text suited for customer service chatbots, search engines, and other purposes. The model comes in three sizes to suit the needs of Nvidia’s clients, with 8 billion, 43 billion, or 530 billion parameters.

“Generative AI is driving the fast adoption of AI and reinventing countless industries,” Nvidia CEO Jensen Huang said. “Nvidia AI Foundations let enterprises customize foundation models with their own data to generate humanity’s most valuable resources — intelligence and creativity.”

On the visual side of generative AI, Picasso allows companies to build and train their own engine to turn text into synthetic images, videos, and 3D objects. Companies can use pre-trained versions of Nvidia’s Edify foundation models or apply their own data to train a custom version of the engine. There are several big brands already working with Nvidia and its new cloud services. Adobe has signed on to create generative AI models to sell on Photoshop, Premiere Pro, and After Effects. Getty Images has also partnered with Nvidia, specifically to create text-to-image and text-to-video models that are “responsibly” sourced from the Getty Images library and with royalties distributed to those who have contributed to training the models. Stock image giant Shutterstock is also announcing a deal with Nvidia despite being a Getty competitor. Shutterstock is more interested in generating 3D objects from text prompts, training the engine on its assets, and similarly compensating the artists whose work helps train the AI.

“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,” Shutterstock CEO Paul Hennessy said. “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.”

Nvidia Deploys Deepfake Eye Contact Feature for Video Streaming

Google Debuts PaLM API and MakerSuite for Generative AI App Developers

Microsoft 365 Copilot Infuses Generative AI Into Every Office App