Elastic Strengthens Production-Scale AI with GPU-Accelerated Vector Search Integrations

Elastic + NVIDIA: GPU-Accelerated Vector Indexing (technical preview)

Elastic and NVIDIA are delivering GPU-accelerated vector indexing in Elasticsearch, powered by NVIDIA cuVS. Elastic provides the real-time context retrieval layer that brings the right data to AI models so they respond and act correctly, while NVIDIA cuVS and NVIDIA GPUs accelerate the heavy compute required to process large vector datasets. This combination enables enterprises to build production-scale AI applications faster and more efficiently. Elasticsearch is also integrated into the NVIDIA Enterprise AI Factory validated design, a full-stack solution for enterprises to build and deploy their own on-premises AI factory.

Elasticsearch powered by NVIDIA cuVS can now index vector data up to 12x faster than CPU-based approaches, removing a major bottleneck that prevents many organizations from deploying generative AI, RAG, and agentic AI systems at production scale.

"As enterprises move from AI experimentation to full-scale production, they require a new class of accelerated data infrastructure to efficiently process massive vector data volumes," said Pat Lee, vice president, Strategic Enterprise Partnerships at NVIDIA. "Our collaboration with Elastic on GPU-accelerated vector indexing, powered by NVIDIA cuVS, provides the high-performance foundation for enterprise deployment requirements for real-time performance at scale for their RAG and agentic AI systems."

This GPU-accelerated vector indexing from Elastic and NVIDIA is also being integrated into new enterprise AI solutions with Dell and Red Hat.

Elastic + Dell: The Dell Data Search Engine

Elastic and Dell developed the Dell Data Search Engine, a vector database powered by Elasticsearch and accelerated by NVIDIA cuVS, natively integrated within the Dell AI Data Platform with NVIDIA.

Built into the Dell AI Data Platform with NVIDIA, the solution enables enterprises to transform petabytes of unstructured data into real-time, actionable context for AI workloads. It allows AI agents to reason, learn, and take action using enterprise data while reducing deployment complexity and overhead costs.

Elastic + Red Hat: AI Factory for Hybrid Cloud

Elastic and Red Hat have collaborated to bring GPU-accelerated vector search powered by NVIDIA cuVS to enterprise hybrid cloud environments. By combining Elastic's real-time search and vector database capabilities with Red Hat OpenShift's Kubernetes capabilities, users get a security-optimized, scalable AI factory reference design for building production-grade RAG systems and autonomous AI agents. Red Hat AI Factory with NVIDIA extends this reference design into a co‑engineered stack that standardizes how these GPU-accelerated RAG and agent workloads are developed, deployed, and managed across hybrid cloud environments.

"Red Hat OpenShift provides the essential, Kubernetes-native foundation for enterprises to operationalize and scale their AI workloads across any hybrid cloud environment," said Katie Giglio, senior director, Ecosystem Development at Red Hat. "By enabling Elastic's GPU-accelerated search on Red Hat OpenShift and Red Hat AI with the speed of NVIDIA, we are jointly delivering a production-ready, open platform that empowers customers to build secure, high-performance RAG and autonomous AI agents while maintaining complete control over their data sovereignty."

"Enterprises are moving rapidly from AI experimentation to full-scale production, and that shift demands a modern data and infrastructure foundation," said Steve Kearns, general manager, Search and AI at Elastic. "With NVIDIA, Dell, and Red Hat, we're enabling organizations to turn massive volumes of unstructured data into real-time context for RAG, generative AI, and agentic systems more efficiently and at enterprise scale."

Together, these integrations expand the Elastic AI Ecosystem, giving organizations flexible options to deploy GPU-accelerated vector search across cloud, hybrid, and on-premises environments.

Availability

• The NVIDIA cuVS integration is currently in technical preview for Elastic self-managed enterprise customers (version 9.3). General availability is scheduled for April 2026 with the version 9.4 release, which will provide a production-ready foundation for high-volume vector search and real-time context retrieval.
• The Dell Data Search Engine is available now with general availability of GPU acceleration targeted for Q2 2026.
• Elasticsearch with GPU acceleration is available on Red Hat AI Enterprise and through the Red Hat AI Factory with NVIDIA.

Additional Materials

• Powering enterprise AI at scale: The Elastic and NVIDIA cuVS integration
• Elastic and Dell AI Data Platform: The foundation for high-velocity enterprise search
• Elastic and Red Hat: Scaling the sovereign AI factory with NVIDIA GPU acceleration

About Elastic

Elastic (NYSE: ESTC), the Search AI Company, integrates its deep expertise in search technology with artificial intelligence to help everyone transform all of their data into answers, actions, and outcomes. Elastic's Search AI Platform — the foundation for its search, observability, and security solutions — is used by thousands of companies, including more than 50% of the Fortune 500. Learn more at elastic.co.

Elastic and associated marks are trademarks or registered trademarks of elasticsearch BV and its subsidiaries. All other company and product names may be trademarks of their respective owners.