At the Computex trade show in Taiwan, Jensen Huang, CEO of Nvidia, delivered a keynote speech highlighting the transformative power of Nvidia NIM (Nvidia inference microservices) in deploying AI applications quickly and efficiently. With NIM, AI models can be deployed within minutes instead of weeks, enabling developers to build generative AI applications for copilots, chatbots, and more.
Nvidia NIM: A Game-Changer for Developers and Enterprises
Nvidia NIM provides developers with optimized containers that serve as inference microservices, allowing them to easily deploy AI models on various platforms such as clouds, data centers, and workstations. This newfound capability significantly increases developer productivity and enables enterprises to maximize their infrastructure investments.
The Growing Complexity of Generative AI Applications
Generative AI applications are becoming increasingly complex, often utilizing multiple models with different capabilities to generate text, images, videos, speech, and more. To address this complexity, Nvidia NIM offers a standardized and simplified approach to incorporating generative AI into applications. This empowers developers to build innovative solutions without the need for a dedicated team of AI researchers.
Expanding Partnerships for Faster Deployment
Over 200 technology partners, including Cadence, Cloudera, Cohesity, DataStax, NetApp, Scale AI, Synopsys, and Hugging Face, have integrated NIM into their platforms to accelerate generative AI deployments. Hugging Face, for example, is now offering NIM with its Meta Llama 3 model. This collaborative effort aims to make generative AI accessible to every organization by integrating it into various platforms and making it widely available to developers.
Applications of NIM in Healthcare and Beyond
NIM’s potential extends beyond gaming and entertainment. Dozens of healthcare companies are leveraging NIM for generative AI inference in surgical planning, digital assistants, drug discovery, clinical trial optimization, and more. NIM’s pre-built containers for GPU-accelerated inference make it easier to deploy applications for generating text, images, video, speech, and digital humans. Researchers can even use NIM microservices for digital biology to accelerate drug discovery by building novel protein structures.
NIM Integration and Deployment Options
Nvidia has partnered with industry leaders and ecosystem partners to embed NIM into their platforms. Platform providers like Canonical, Red Hat, Nutanix, and VMware are supporting NIM on open-source KServe or enterprise solutions. Leading AI tools and MLOps partners such as Amazon SageMaker, Microsoft Azure AI, and DataRobot have also integrated NIM into their platforms. Additionally, global system integrators and service delivery partners like Accenture and Deloitte have created NIM competencies to facilitate the rapid development and deployment of production AI strategies.
NIM-enabled Applications Everywhere
Enterprises can deploy NIM-enabled applications virtually anywhere, including Nvidia-certified systems from global infrastructure manufacturers like Cisco, Dell Technologies, Hewlett-Packard Enterprise, Lenovo, Supermicro, and server manufacturers ASRock Rack, Asus, Gigabyte, Ingrasys, Inventec, Pegatron, QCT, Wistron, and Wiwynn. Furthermore, NIM microservices have been integrated into major cloud platforms such as Amazon Web Services, Google Cloud, Azure, and Oracle Cloud Infrastructure.
Real-World Applications of NIM
Several industry leaders, including Foxconn, Pegatron, Amdocs, Lowe’s, and ServiceNow, have embraced NIM for generative AI applications in manufacturing, healthcare, financial services, retail, customer service, and more. Foxconn, the world’s largest electronics manufacturer, is utilizing NIM to develop domain-specific LLMs for smart manufacturing, smart cities, and smart electric vehicles.
Nvidia-Certified Systems for AI and Accelerated Computing
Nvidia is expanding its Nvidia-certified systems program, certifying partner systems for AI and accelerated computing. The program now includes Spectrum-X Ready systems for AI in the data center and IGX systems for AI at the edge. These certified systems undergo rigorous testing to ensure enterprise-grade performance, manageability, security, and scalability for Nvidia AI workloads.
Facilitating Deployment with Nvidia NIM and KServe
Nvidia NIM works seamlessly with KServe, an open-source software that automates the deployment of AI models at scale. This integration enables developers, ecosystem partners, and customers to access the performance, support, and security of the Nvidia AI Enterprise software platform through NIM. With an API call, generative AI can be deployed like any other enterprise application, making it easier than ever to incorporate AI into various domains.
Accelerating Healthcare and Life Sciences Workflows with Llama 3
Meta Llama 3, a state-of-the-art large language model optimized using Nvidia accelerated computing, is now available as a downloadable NIM inference microservice. Healthcare developers, researchers, and companies can leverage Llama 3 to innovate responsibly in applications ranging from surgical planning to drug discovery. The standardized API allows for easy deployment anywhere.
In conclusion, Nvidia NIM is revolutionizing the deployment of generative AI models by providing developers with optimized containers and standardized microservices. This technology is empowering enterprises across industries to unlock the full potential of AI and accelerate their digital transformation. With NIM’s integration into various platforms and partnerships with industry leaders, the future of generative AI looks promising.
