IBM and NVIDIA join forces to accelerate enterprise AI adoption
IBM and NVIDIA have joined forces to create enterprise-level AI solutions and promote the large-scale application of artificial intelligence.
Core content:
1. IBM and NVIDIA work together to integrate products based on the AI data platform reference architecture
2. IBM Fusion adds content-aware storage capabilities to improve AI performance and ease of use
3. IBM Cloud provides use cases based on NVIDIA H200 to enhance support for compute-intensive workloads
Recently, IBM (NYSE: IBM) announced a new collaboration with NVIDIA (NASDAQ: NVDA). The two parties will integrate products based on the NVIDIA AI data platform reference architecture to help enterprises more effectively use data to build, expand and manage generative artificial intelligence (Gen AI) workloads and AI intelligent applications . In addition, IBM will add content-aware storage capabilities to its hybrid cloud infrastructure product IBM Fusion and expand the integration of the NVIDIA platform with IBM's AI and data platform WatsonX . In addition, IBM will work with NVIDIA to launch a new IBM Consulting solution to promote enterprise-level AI innovation.
According to a 2024 survey by the IBM Institute for Business Value, more than three-quarters (77%) of executives surveyed said generative AI is ready to enter the market, up from 36% in 2023. While pushing AI into production, enterprises are also increasing their demand for computing and data-intensive technologies . The collaboration with NVIDIA will support IBM in providing hybrid AI solutions based on open technologies and platforms, while taking into account data management, performance, security and governance.
The following solutions are based on the NVIDIA AI Data Platform reference architecture and represent the latest results of the collaboration between the two parties to build enterprise-grade AI infrastructure:
Enhanced unstructured data processing and improved AI performance : With IBM's new content-aware storage (CAS) capabilities, enterprises can extract insights and use them for reasoning from rapidly growing unstructured data without compromising trust and security, and responsibly expand and enhance application scenarios such as "retrieval augmentation generation" (RAG) and AI reasoning. IBM Storage Scale distributed storage software will use extracted and enhanced data to respond to queries and leverage NVIDIA BlueField-3 DPUs and NVIDIA Spectrum-X networks to speed up communication between GPUs and storage. Multimodal document data extraction workflows will also use the NVIDIA NeMo Retriever microservice built on the NVIDIA NIM platform. CAS will be embedded in the next update of IBM Fusion, which is expected to be released in the second quarter of this year.
Making AI more accessible : IBM WatsonX will integrate with the NVIDIA NIM platform to help enterprises use industry-leading AI models in multi-cloud environments. Enterprises can use IBM's enterprise-grade AI platform and developer studio watsonx.ai to develop AI models and deploy them to their chosen applications while using other external models. IBM's AI governance platform watsonx.governance ensures that enterprises can implement strong monitoring and governance of NVIDIA NIM microservices in any hosting environment. As more enterprises adopt AI agents and other advanced applications that integrate AI models, interoperability between systems becomes increasingly important.
Enhanced support for compute-intensive workloads : As more enterprises adopt generative AI and high-performance computing (HPC), IBM continues to expand its NVIDIA-powered computing portfolio by announcing the availability of NVIDIA H200-based use cases on IBM Cloud. With large memory and high bandwidth, these use cases meet enterprise demands for AI workloads and large underlying models.
Partnering with NVIDIA to Infuse AI Agents into Process Transformation : IBM Consulting will launch AI integration services to help clients use NVIDIA Blueprints and AI agent technologies to transform and manage end-to-end business processes, such as applying AI agents to industry-specific workflows in edge environments. Use cases include autonomous inspection and maintenance in manufacturing, and proactive analysis of video data and anomaly response in the energy industry.
Optimizing compute-intensive AI workloads in hybrid cloud environments : IBM Consulting helps clients build, update, and manage compute-intensive AI workloads in hybrid cloud environments with RedHat OpenShift and NVIDIA AI. This includes technologies such as NVIDIA AI Foundry, NVIDIA NeMo, NVIDIA AI Enterprise, NVIDIA Blueprints, and NVIDIA Clara to accelerate complex, compute-intensive tasks while meeting AI governance, data security, and compliance requirements.
“IBM is committed to helping enterprises build and deploy effective AI models and quickly scale them,” said Hillery Hunter, CTO and general manager of IBM’s Infrastructure Business Innovation Group . “IBM and NVIDIA are working together to create the solutions, services, and technologies necessary to help clients unlock, accelerate, and protect the use of data, ultimately reducing the hidden costs and technical barriers to AI and gaining real business benefits.”
“AI agents need to access, acquire and process data quickly and at scale, and today, these steps occur in separate silos,” said Rob Davis, vice president of storage networking technology business at NVIDIA . “The integration of IBM’s content-aware storage and NVIDIA AI helps enterprises achieve near real-time inference and responsive AI inference by coordinating data and computing resources through an optimized network fabric, eliminating technology silos through intelligent, scalable systems.”
