What Is NVIDIA NIMs ? How Its Revolutionizing Generative AI Deployment?

At COMPUTEX 2024, NVIDIA's founder and CEO, Jensen Huang, reveled a New advancement in AI model deployment: NVIDIA NIMs. These microservices are designed to simplify and expedite the integration of AI into various applications, significantly reducing deployment times from weeks to mere minutes. This blog explores how NVIDIA NIMs are set to revolutionize the landscape of generative AI, making it more accessible and efficient for enterprises worldwide.

Addressing the Complexity of Generative AI

The surge in demand for complex AI and generative AI (GenAI) applications has highlighted the need for a streamlined deployment process. Tasks such as text, image, video, and speech generation require multiple models, creating a complex environment for developers. NVIDIA NIMs provide a standardized approach to embedding AI into applications, enhancing developer productivity and enabling enterprises to maximize their existing infrastructure investments.

For example, running Meta Llama 3-8B on a NIM system generates up to three times more AI tokens on accelerated infrastructure compared to non-NIM systems. This demonstrates the efficiency and resource optimization that NIMs bring to AI deployment.

Key Features of NVIDIA NIMs

Speed and Efficiency

NVIDIA NIM microservices are pre-built to speed up model deployment for GPU-accelerated inference. They incorporate NVIDIA's powerful software tools, including CUDA, Triton Inference Server, and TensorRT-LLM, to ensure optimal performance. Over 40 models from the NVIDIA community are available as NIM endpoints, providing developers with easy access to a wide range of resources.

Accessibility and Integration

One of the standout features of NVIDIA NIMs is their accessibility. Developers can now deploy and execute Llama 3 NIMs via the Hugging Face AI platform with just a few clicks, using NVIDIA GPUs on their preferred cloud infrastructure. This ease of use extends to various generative tasks, such as text, image, video, speech, and digital human generation.

Broad Partner Support

NVIDIA NIMs have garnered extensive support from over 150 technology partners. Companies like Cadence, Cloudera, Cohesity, DataStax, NetApp, Scale AI, and Synopsys are integrating NIMs into their platforms to expedite generative AI deployment in domain-specific applications. AI infrastructure partners, including Canonical, Red Hat, Nutanix, VMware, Amazon SageMaker, Microsoft Azure AI, and Dataiku, are also embedding NIMs into their platforms, enabling developers to build and deploy domain-specific AI applications with optimized inference.

Application in Various Industries

Numerous industries are already leveraging NVIDIA NIMs for generative AI applications. In healthcare, companies are using NIMs to enhance surgical planning, digital assistants, drug discovery, and clinical trial optimization. The ACE NIM is available for developers to create lifelike digital humans for customer service, telehealth, education, gaming, and entertainment.

Customer Success Stories

Several corporations have successfully implemented NVIDIA NIMs to enhance their operations:

• Foxconn is developing domain-specific large language models (LLMs) for smart manufacturing, smart cities, and smart electric vehicles using NIMs.

• Pegatron is utilizing NIMs for Project TaME, a local LLM development initiative.

• Amdocs has enhanced its customer billing LLM, achieving significant improvements in cost, accuracy, and latency.

• Lowes is employing NIM microservices to improve customer and associate experiences with generative AI.

• ServiceNow is integrating NIM within its Now AI multimodal model to facilitate scalable and cost-effective LLM development.

• Siemens is using NIM microservices for shop floor AI workloads and building an on-premises Industrial Copilot for Machine Operators.

Conclusion

NVIDIA NIMs represent a significant leap forward in the deployment of generative AI. By simplifying and accelerating the integration of AI into various applications, NIMs are making it possible for enterprises of all sizes to Used the power of AI without the need for extensive AI research teams. As more industries adopt this technology, the potential for innovation and efficiency gains is immense. NVIDIA NIMs are truly democratizing access to generative AI, placing it within reach for every organization.