NVIDIA Introduces AI Foundation, Generative AI Cloud Service That Can Be Adjusted

Photo: NVIDIA AI Foundation (Photo: NVIDIA)

JAKARTA - During the 2023 GPU Technology Conference (GTC), NVIDIA announced three cloud services named the NVIDIA AI Foundation, which allows companies to build and run a special large language model and a trained generative AI model for specific tasks.

The three cloud services are NVIDIA NeMo, a text-to-image generation machine.

Next up is NVIDIA BioNemo, a NeMo model focused on the discovery of drugs and molecules created for the medical research community, and finally Picasso, an AI model capable of producing images, videos, and 3D applications to increase productivity of creativity, design, and digital simulations.

Both NeMo and Picasso services operate on the new NVIDIA DGX Cloud platform and will eventually be accessible via online portals, although currently still the early stages of development.

"Generative AI encourages rapid adoption of AI and recreate countless industries. The NVIDIA AI Foundation allows companies to adapt the foundation model to their own data to produce the most valuable human resources of intelligence and creativity," said Jensen Huang, founder and CEO of NVIDIA.

Just launched at #GTC23, NVIDIA AI Foundations, a set of cloud services that provides businesses with a simplified approach to build, customize, and run #generativeAI models with their proprietary data for text language, visual content, and biology. https://t.co/qq0WlULZpS pic.twitter.com/bwi8yag2ur

— NVIDIA AI (@NVIDIAAI) March 21, 2023

According to NVIDIA, each cloud service includes six elements: pre-training models, frameworks for data processing, databases and vector personalization, optimized inference machines, APIs, and support from NVIDIA experts to help companies refine models for their special use cases.

This cloud-based service serves as an empty template that companies can use to store their own databases for specialized training.

So the estimate is like this, if Google's Bard AI is trained and withdraws data from all the information on the internet to respond to users, NVIDIA AI will allow companies to customize their own Big Language Model (LLM) according to their specific needs.

Thus, the AI NVIDIA model allows companies to customize models to support accurate generative AI applications for market intelligence, company search, chatbots and customer service, and more.

Among the big companies, Getty Images, Morningstar, Quantiphi, and Shutterstock are some companies that will use the NVIDIA AI Foundation to create and use AI models, applications that include languages, images, videos, and 3D.