Enhancing search performance at scale using AWS Graviton with Elastic

Founded in 2012, Elastic, an AWS Partner, has grown from an open-source search engine project into a comprehensive data solution that serves over 21,000 customers worldwide—with a continually expanding customer base. The company has built a search AI data lake that delivers results in milliseconds, while remaining cost efficient. Through its partnership with AWS, Elastic is shaping the future of innovation, driving leadership across observability, security, and generative AI. This market leadership is validated by recognition as a Leader in both the Forrester Wave: Security Analytics Platforms, Q2 2025, and the 2025 Gartner Magic Quadrant for Observability Platforms. These industry accolades highlight Elastic’s ability to innovate and deliver value at scale.

The company’s Elasticsearch technology powers critical search, observability, and security solutions that help organizations analyze terabytes of data daily and respond to threats in near real time (NRT). The company plays a vital role in the AI space, supporting almost everything from enterprise search applications to machine learning workloads. Elastic combines full-text and vector search to power generative AI applications including AI assistants, semantic search, Retrieval Augmented Generation systems, and agentic applications with robust security and scalability.

However, supporting diverse workloads at such a massive scale presented significant infrastructure challenges. Elastic’s search operations are inherently CPU intensive. They require substantial computational power for complex queries, relevance scoring and ranking algorithms, and NRT document indexing—all while maintaining millisecond response times. The company needed a solution that could deliver high performance for these demanding CPU workloads while maintaining cost efficiency for storage-heavy security operations. This led Elastic to explore new strategies for optimizing its compute infrastructure on AWS. “We engaged with AWS because it matches our pace of innovation and supports our scaling needs,” says Udayasimha Theepireddy, director of cloud architecture at Elastic.

After evaluating various options, Elastic found the answer in AWS Graviton processors: custom-designed server processors developed by AWS to provide excellent price performance for cloud workloads running on Amazon Elastic Compute Cloud (Amazon EC2)—which provides secure and resizable compute capacity for virtually any workload. AWS delivers the robust compute technology and cost-effective infrastructure that Elastic needs to scale and innovate at the speed of its customers’ demands.

To identify a solution for addressing its infrastructure needs, Elastic’s performance benchmarking team conducted extensive internal testing and validation across different AWS Graviton–based Amazon EC2 instance families. Ultimately, the team chose Amazon EC2 C6g Instances, compute-optimized instances for applications that need high-speed, low-latency local storage. The team also adopted Amazon EC2 C7g Instances, which are designed for compute-intensive workloads.

Unlike traditional virtual processors, each virtual CPU on AWS Graviton–based instances is backed by a physical processor core. This architectural advantage proved particularly beneficial for Elastic’s diverse workload requirements, spanning almost all workloads and use cases. What’s more, Graviton processors use up to 60 percent less energy than comparable instances, helping to optimize price performance.

For new-to-Elastic customers, AWS Graviton is the default deployment option, requiring no additional configuration or technical expertise. Customers can create a deployment, and their Elasticsearch observability, security, or search workloads automatically run on the optimized AWS Graviton infrastructure. To facilitate adoption among existing customers, Elastic developed a single-click migration process to handle all data transfer automatically with virtually no downtime. This feature removes the technical barriers that might otherwise prevent existing customers from using AWS Graviton.

Elastic recommends AWS Graviton to customers across all deployment options, including its self-managed Elasticsearch solutions through Elastic Cloud Enterprise and Elastic Cloud on Kubernetes. AWS Graviton–based instances can be especially helpful for organizations in the public sector and other regulated industries where self-managed deployments are preferred. AWS Graviton also powers most of Elastic Cloud Serverless, delivering excellent price performance and the same infrastructure optimizations across the company’s entire product portfolio.

“Elastic improved its workloads’ price performance by up to 35 percent compared with similar instances,” says Yuvraj Gupta, principal product manager at Elastic. This solution improved indexing throughput for write-intensive workloads and reduced search latency for read-heavy operations.

“Our customers can improve price performance by using Elastic on AWS Graviton instances and invest these savings in building new and innovative applications,” says Theepireddy. This creates a virtuous cycle where customers bring more data and use cases to Elastic and benefit from new features.

Elastic is now accelerating its investment in generative AI to meet the growing demand for AI-powered search and analytics solutions. Using AWS compute technology, combined with comprehensive capacity planning and technical support, Elastic has the reliable foundation it needs to serve a large customer base while maintaining millisecond response times.

Now, the company can focus on what it excels at: developing cutting-edge features to help customers unlock insights from their data without worrying about infrastructure limitations. “Using AWS, we are giving that innovation and entrepreneurial spirit to our customers,” says Theepireddy. “They can build generative AI solutions cost effectively.”