NVIDIA today launched its customizable Gen-AI cloud service called AI Foundations. This cloud-based service will allow businesses to create, refine, and operate custom large language models and generative AI models. These models can be trained with their own proprietary data and created for their unique domain-specific tasks. With this new offering, NVIDIA aims to enable businesses that lack the time and money to develop their own models to leverage the power of AI.

The race to dominate the enterprise AI market is in full swing, with major tech companies such as Google, Microsoft, Meta, and Baidu competing to incorporate generative AI features into their existing products. While some companies like Adobe and AT&T are scrambling to find uses for the revolutionary technology, heavyweights like Google and Microsoft are jockeying their Large Language Models (LLMs) for market dominance.

NVIDIA’s AI Foundations offers a range of models, including NeMo, NVIDIA’s text-to-image generation engine, BioNemo, a drug and molecule discovery-focused fork of the NeMo model built for the medical research community, and Picasso, an AI capable of generating images, video, and 3D applications to supercharge productivity for creativity, design, and digital simulation. Although NeMo and Picasso are still in early access and private preview, respectively, they will eventually be accessible through an online portal.

Unlike some AI models, NVIDIA’s models allow companies to tailor an LLM to their own specific needs using their own proprietary data. The models can be trained with up to 530 billion parameters, which is more than triple the 185 billion parameters provided by GPT-3.5. For example, NVIDIA announced a system built on the NeMo cloud service that is trained on Getty Images‘ fully licensed assets. The system will provide royalties to artists on any revenues generated from the models.

BioNeMo, which uses the same technical underpinnings as NeMo, is geared entirely toward drug and molecule discovery. The service enables researchers to fine-tune generative AI applications on their own proprietary data and run AI model inference directly in a web browser or through new cloud APIs that easily integrate into existing applications (read more of this post).

NVIDIA has also announced partnerships with Adobe and Shutterstock. As part of Adobe’s Content Authenticity Initiative, NVIDIA will incorporate many of Picasso’s features directly into Adobe’s suite of editing software, including Photoshop, Premiere Pro, and After Effects (read more of this post). Shutterstock will use Picasso to generate 3D objects from text prompts as a new feature within Creative Flow, with plans to offer it on Turbosquid.com and NVIDIA’s forthcoming Omniverse platform.

NVIDIA’s AI Foundations offering is a game-changer in the enterprise AI market, providing businesses with the power to create their own AI models and incorporate generative AI features into their products. With the ability to train models with their own proprietary data, companies can tailor AI models to their unique needs and gain a competitive edge in the market.