OpenSearch Realizes Impressive Performance and Performance/Dollar Gains on AWS I7i VMs on Intel®

Authors:

Isaac Gonzalez: Cloud Software Engineer

Peter Velasquez: Sales Account Manager

Introduction

Search and analytics workload requirements are rapidly evolving as business needs change. Recent search and analytics trends include:

• AI-enhanced search as users expect to “understand meaning” not just keywords.
• Real-time analytics and “instant insight.”
• Cost Optimization driving demand for open-source solutions.

OpenSearch is a leading open-source search and analytics suite for developers, data scientists, and enterprises seeking fast, large-scale search to meet these evolving needs. Its robust aggregation and search capabilities, horizontal scalability, and real-time analytics make it highly valuable across numerous industries and applications.

AWS EC2 I7i VMs powered by 5th Gen Intel® Xeon® Processors Take on the Performance Challenge

Amazon’s (AWS) EC2 I7i instances, powered by 5th Gen Intel® Xeon® processors (formerly codenamed Emerald Rapids), introduce quantifiable improvements in compute architecture, memory subsystem, and I/O bandwidth, extending the storage-optimized lineage of Intel-powered instances on AWS. Building on the AWS I4i generation, the I7i family expanded L3 cache and DDR5 memory, delivering greater bandwidth, all contributing to measurable performance gains for data-intensive workloads like OpenSearch.

What's New with I7i

The I7i family introduces several architectural and platform-level upgrades over I4i, grounded in documented specifications from AWS:

• 5th Gen Intel® Xeon® processors
• All-core turbo frequency up to 3.2 GHz (maximum turbo 4.0 GHz)
• Expanded L3 cache for improved data locality in analytics workloads
• DDR5 memory support, providing higher memory bandwidth than DDR4
• PCIe 5.0 connectivity and faster NVMe SSDs for higher I/O throughput

These enhancements contribute to greater efficiency in memory-intensive, I/O-heavy analytics, and search workloads. This translates into higher query throughput and better total cost of ownership (TCO) for search and analytics clusters.

Our analysis evaluated how these architectural enhancements impact OpenSearch histogram aggregation performance and performance-per-dollar efficiency across identical cluster configurations. We compared the I7i and I4i instance families across multiple instance sizes under consistent test conditions.

What We Found

Preliminary results show that I7i instances deliver up to 1.75× increased throughput with consistent gains across all sizes. (See table 1 in endnotes)

OpenSearch Throughput Performance on I7i

 Figure 1: OpenSearch Throughput Comparison (I7i vs I4i)

Impressive Performance per Dollar Results

Beyond raw performance, the I7i family also demonstrated improved performance per dollar, with a throughput-per-search-cost increase of up to 59%, driven by architectural efficiency and competitive pricing. (See table 2 in Endnotes)

Figure 2: Performance-per-Dollar Comparison (I7i vs I4i)

(Ops/s / $/s )/1000= 1000 Ops per dollar

These results underscore that customers upgrading from AWS I4i to I7i VM instances can expect faster query completion times and lower per-query costs.

Why it Matters

For OpenSearch users, higher throughput and efficiency translate directly into improved responsiveness, lower infrastructure footprints, and better ROI.

• Developers can support more concurrent users or queries without scaling horizontally.
• Data engineers can achieve higher query throughput and lower latency under the same workload.
• Businesses can reduce their cloud spend while maintaining or improving SLAs.

Conclusion

Powered by 5th Gen Intel® Xeon® processors, the AWS I7i family delivers a compelling leap forward in performance gains and cost efficiency for storage-optimized workloads.

Compared to previous gen I4i, I7i offers higher throughput, improved performance-per-dollar, and stronger scaling across all tested sizes.

For organizations running OpenSearch or other I/O-intensive services, migrating to AWS I7i can yield measurable performance gains while maintaining the flexibility and reliability of Intel-powered AWS infrastructure.

Endnotes and Configuration Data

Benchmark Setup

The same methodology and OpenSearch Benchmark tooling used in our previous I4i vs I3 comparison were applied here. To ensure consistency, we normalized throughput values using the I4i instances as the baseline for each size category.

• Workload: nyc_taxis dataset: 165 million documents, 75 GB of data
• Metrics:
  • Histogram aggregation throughput (operations per second)
  • Performance per dollar (throughput / instance hour cost)
• Cluster configuration:
  • 3 data nodes, 1 coordinating node, 1 manager node
  • Benchmark client node running OpenSearch Benchmark 1.1
  • JVM heap = 50% of RAM
  • Translog threshold = ¼ heap size
  • Index buffer = 25% heap

All tests ran in the same AWS region using on-demand pricing, with storage and networking held constant.

Instance sizes evaluated: 2xlarge, 4xlarge, and 8xlarge.

Table 1: Performance Data

Table 2 Performance/Dollar Data

Notices and Disclaimers

Performance varies by use, configuration, and other factors. Learn more on the Performance Index site.
Performance results are based on testing as of dates shown in configurations and may not reflect all publicly available ​updates. See backup for configuration details. No product or component can be absolutely secure.
Your costs and results may vary.
Intel technologies may require enabled hardware, software, or service activation.
© Intel Corporation. Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries. Other names and brands may be claimed as the property of others.