NVIDIA and Dell recently partnered to launch a product called Project Helix, which aims to introduce the power of generative AI to enterprises. Providing an easy way for enterprises to get started with generative AI, enabling them to use their own data to build new or customize existing generative AI base models. Because this work is done within the enterprise entity, Project Helix can also help mitigate the intellectual property leakage concerns that many companies have.
▲ Image Source NVIDIA Newsroom
The product focuses on creating full-stack, on-premises generative AI solutions that enable companies to build new or customize existing generative AI base models using their own data. Because the product is specifically designed to simplify the process of getting started with generative AI, it helps organizations avoid common mistakes and cost overruns when implementing generative AI projects.
Project Helix provides organizations with AI models specifically built to get value out of today’s massive amounts of underutilized data faster and more securely,” said Jeff Clarke, vice chairman and co-chief operating officer, Dell Technologies Group. “With a highly scalable and efficient infrastructure, enterprises can create a new wave of generative AI solutions that will reshape their industries.
As a company closely associated with AI, at NVIDIA’s last GTC conference in March, they unveiled a full suite of generative AI-related software, including industry-specific software foundation models and enterprise-centric development tools, specifically its NeMo Large Language Model (LLM) framework and NeMo Guardrails for filtering unwanted topics, which surprised many people. These models are optimized to run on NVIDIA hardware.
The Helix project also represents a collaborative effort between NVIDIA and Dell, with Dell bundling NVIDIA’s enterprise AI software through a series of PowerEdge server systems. In addition, Dell offers several different storage options from its PowerScale and ECS enterprise object storage product lines that are optimized for AI workloads. The result is a comprehensive solution that enables enterprises to begin building or customizing generative AI models. Potential customers can use NVIDIA’s base model option, or they can select an open source model from Hugging Face (or other technology vendor solutions) and begin the process.”
It is revealed that Project Helix products will be available through traditional channels and APEX flexible consumer options starting in July 2023.