- NVIDIA AI Foundry allows businesses to create and deploy custom generative AI models using NVIDIA’s infrastructure and tools.
- The service includes access to NVIDIA DGX Cloud, foundation models, NVIDIA NeMo software, and expert support.
- Companies can customize various models, including Llama 3.1, NVIDIA Nemotron, and others from the open community.
- Prominent firms such as Amdocs, Capital One, Getty Images, KT, Hyundai Motor Company, SAP, ServiceNow, and Snowflake are early adopters.
- The service provides support through a global ecosystem of partners, including Accenture, Deloitte, Infosys, and Wipro.
- Models can be deployed as NVIDIA NIM inference microservices for optimized performance.
- Together AI will utilize NVIDIA’s infrastructure for deploying models, enhancing performance and scalability.
- NVIDIA NeMo tools streamline data curation, model fine-tuning, and performance evaluation.
Main AI News:
NVIDIA has launched AI Foundry, a new service designed to help enterprises create tailored generative AI models that meet their specific industry needs. This innovative offering allows businesses to leverage NVIDIA’s advanced computing infrastructure and software tools to develop and deploy custom AI models, enhancing their AI initiatives.
Similar to how TSMC manufactures chips based on other companies’ designs, NVIDIA AI Foundry provides the infrastructure for businesses to build and refine their AI models. Using resources such as DGX Cloud, NVIDIA NeMo software, and various foundation models, AI Foundry supports companies in crafting models that fit their particular requirements. Unlike TSMC’s physical chips, NVIDIA AI Foundry focuses on generating digital models, facilitating innovation through a comprehensive ecosystem of tools and expertise.
Enterprises can customize a wide range of models through AI Foundry, including NVIDIA’s own offerings and models from the open community such as Llama 3.1, NVIDIA Nemotron, CodeGemma by Google DeepMind, and others. This customization capability is essential for businesses aiming to integrate AI into their workflows efficiently.
Prominent companies including Amdocs, Capital One, Getty Images, KT, Hyundai Motor Company, SAP, ServiceNow, and Snowflake are among the first to utilize NVIDIA AI Foundry. These pioneers are leveraging the service to drive advancements in software, technology, communications, and media sectors. Jeremy Barnes, Vice President of AI Product at ServiceNow, noted, “Organizations deploying AI can gain a competitive edge with custom models that incorporate industry and business knowledge.”
NVIDIA AI Foundry is supported by several key components: foundation models, enterprise software, accelerated computing, expert support, and a broad partner ecosystem. The software suite includes NVIDIA’s AI foundation models and the NVIDIA NeMo platform, which streamlines model development. DGX Cloud, a network of accelerated computing resources in partnership with leading public cloud providers like AWS, Google Cloud, and Oracle Cloud, enables seamless model development and scaling.
For additional support, NVIDIA AI Enterprise experts are available to assist customers through the model building, fine-tuning, and deployment processes. The service also connects users with a global network of partners such as Accenture, Deloitte, Infosys, and Wipro, which offer consulting and implementation services. Accenture has introduced the AI Foundry-based Accenture AI Refinery framework for custom model development.
Service delivery partners like Data Monsters, Quantiphi, Slalom, and SoftServe help businesses integrate AI into their existing IT environments, ensuring scalability and security. NVIDIA’s partners provide AIOps and MLOps platforms for production use, and models can be deployed as NVIDIA NIM inference microservices, optimized for performance and efficiency.
Together AI, a leading AI acceleration cloud, has announced that its ecosystem will use NVIDIA GPU-accelerated inference stacks to deploy models such as Llama 3.1 on DGX Cloud, enhancing performance and scalability.
NVIDIA NeMo further simplifies the custom model development process. Tools such as NeMo Curator, Customizer, Evaluator, and Guardrails support data curation, model fine-tuning, performance evaluation, and application safety, respectively. This integration allows businesses to create models tailored to their needs, improving accuracy and operational efficiency. Philipp Herzig, Chief AI Officer at SAP, highlighted that SAP plans to use NeMo to enhance AI-driven productivity through SAP Business AI.
Conclusion:
NVIDIA AI Foundry is set to significantly impact the market by offering enterprises a powerful platform for developing customized generative AI models. This innovation enables businesses to create AI solutions that are finely tuned to their specific industry needs, thereby improving accuracy and operational efficiency. As companies increasingly seek tailored AI applications, NVIDIA AI Foundry positions itself as a crucial player in the competitive AI landscape, facilitating advanced and scalable solutions that address unique business requirements. This development could drive a shift towards more specialized and effective AI implementations across various sectors.
