Leaders in the Storage and NVIDIA Industries Reveal a New Era of Enterprise Infrastructure for AI Storage Providers Construct Infrastructure with AI Query Agents. That Use NVIDIA Networking, Software, and Computing to Reason and Quickly Produce Precise Answers to Complicated Questions.
In order to meet the demands of demanding AI inference workloads, NVIDIA today announced the NVIDIA AI Data Platform, a customizable reference design that top providers are using to create a new class of AI infrastructure: enterprise storage platforms with AI query agents powered by NVIDIA accelerated computing, networking, and software.
NVIDIA-Certified Storage providers can use the NVIDIA AI Data Platform to provide infrastructure that uses specialised AI query agents to accelerate workloads related to AI reasoning. Using NVIDIA AI Enterprise technologies, such as NVIDIA NIM microservices for the new NVIDIA Llama Nemotron models with reasoning capabilities and the new NVIDIA AI-Q Blueprint, these agents assist businesses in producing insights from data in almost real time.
With NVIDIA Spectrum-X networking, NVIDIA Blackwell GPUs, NVIDIA BlueField DPUs, and the open-source inference library NVIDIA Dynamo, storage providers may optimize their infrastructure to power these agents.
In order to develop specialised AI data platforms that can use enterprise data to reason and answer complex questions, NVIDIA is working with top data platform and storage providers, such as DDN, Dell, HPE, Hitachi Vantara, IBM, NetApp, Nutanix, Pure Storage, VAST Data, and WEKA.
In the era of artificial intelligence, data is the raw material that drives industries. “The nation is developing a new class of enterprise infrastructure with the top storage companies in the world, which businesses need to implement and scale agentic AI across hybrid data centres.”
NVIDIA AI Data Platform Enhances Storage with AI and Accelerated Computing
For the millions of companies that use enterprise storage for the data that powers their operations, the NVIDIA AI Data Platform offers enhanced computing and AI.
AI query agents may access data stored on corporate systems more quickly because to an accelerated engine provided by NVIDIA Blackwell GPUs, BlueField DPUs, and Spectrum-X networking. With up to 50% less power usage and up to 1.6 times greater performance than CPU-based storage, BlueField DPUs offer more than three times the performance per watt. Spectrum-X uses adaptive routing and congestion control to speed up AI storage traffic by up to 48% when compared to standard Ethernet.
The NVIDIA AI-Q Blueprint is used in AI Data Platform storage infrastructure to create agentic systems that can connect to enterprise data and reason. AI-Q uses NVIDIA NeMo Retriever microservices to speed up data retrieval and extraction on NVIDIA GPUs by up to 15 times.
In order to offer more precise, contextually aware answers, AI query agents constructed using the AI-Q Blueprint connect to data during inference. They can swiftly access vast amounts of data and process many kinds of data, including unstructured, semi-structured, and structured data from a variety of sources, such as text, PDFs, photos, and videos.
Storage Industry Leaders Building AI Data Platforms With NVIDIA
- NVIDIA is working with its Certified Storage partners to develop unique AI data platforms.
- DDN’s DDN Infinia AI platform is being designed with AI Data Platform features.
- For its PowerScale and Project Lightning product lines, Dell is developing AI data platforms.
- HPE Private Cloud for AI, HPE Data Fabric, HPE Alletra Storage MP, and HPE GreenLake for File Storage are all getting AI Data Platform capabilities from Hewlett Packard Enterprise.
- By integrating the AI Data Platform into the Hitachi IQ ecosystem, Hitachi Vantara is assisting clients in innovating with data offerings and storage solutions that produce observable AI results.
- To speed up retrieval-augmented generation applications, IBM is combining AI Data Platform with IBM Fusion and IBM Storage Scale technology as part of its content-aware storage capability.
- With the NetApp AIPod solution, which is based on the AI Data Platform, NetApp is improving enterprise storage for agentic AI.
- By integrating with the NVIDIA AI Data Platform, Nutanix Cloud Platform with Nutanix Unified Storage will facilitate inferencing and agentic workflows that are implemented across edge, data centre, and public cloud environments.
- With the Pure Storage FlashBlade, Pure Storage will provide AI Data Platform capabilities.
- VAST Data and AI Data Platform are collaborating to use VAST InsightEngine to curate real-time insights.
- The WEKA Data Platform software provides a high-performance storage foundation that speeds up workloads related to AI inference and token processing by integrating with NVIDIA GPUs, DPUs, and networking to optimize data access for agentic AI reasoning and insights.
Beginning this month, NVIDIA-Certified Storage providers will be able to offer solutions developed using the NVIDIA AI Data platform.
NVIDIA Enterprise AI Data Platform
In order to provide AI agents with near-real-time business insights, the NVIDIA AI Data Platform combines corporate storage with NVIDIA-accelerated compute. It is a configurable reference architecture for a new class of AI infrastructure.
Leaders in business storage developed these data systems to boost agentic AI workflow accuracy and performance. Spectrum-X networking, NVIDIA AI Enterprise software, NVIDIA Blackwell GPUs, and BlueField-3 DPUs are all included in the platform.
Explore the Benefits of the NVIDIA AI Data Platform
Business data is unlocked via faster enterprise storage, which improves AI agents.
High-Performance Computing Infrastructure for Agentic AI
The platform combines network connectivity and AI data processing with the capabilities of accelerated computation. Workflows are optimized, allowing AI agents to analyze and act on massive amounts of business knowledge more quickly and effectively.
Distributed Inference, Built for Enterprise Scale
In order to speed up data processing and enable scalable, sophisticated AI workflows, the platform uses intelligent routing, load balancing, and enhanced caching to optimise the distribution of AI workloads throughout GPUs, nodes, and LLM memory management.
Comprehensive Data Support
Businesses can use the platform to process and extract insights from both structured and unstructured data, gaining access to important information from all available data sources, including text, PDFs, photos, and videos.
Native Support for Retrieval- Augmented Generation
The accuracy of AI agents‘ responses is improved by the platform’s smooth integration of proprietary company knowledge, such as internal papers, customer records, and domain-specific data, into AI workflows.
