Up to 57 percent faster performance delivers quicker insights and lower costs for subsurface energy research.
Today, we reveal that Cloud Native Qumulo (CNQ) on Azure significantly outperformed Azure Files V2 in testing with Petrel subsurface software using the Cegal Geophysics Benchmarking Workflow. The side-by-side evaluation—run on similarly priced and configured Azure-based environments—showed CNQ delivering up to 57 percent faster performance than Azure Files V2.
Cegal is a Qumulo partner and leader in Oil and Gas workload services and data insights. Their benchmarking standards are recognized in the industry for excellence and applicability to real-world use cases. Cegal’s data assessments and tools help clients understand and optimize their IT and subsurface application environments. This includes identifying performance bottlenecks in applications like Petrel.
“We’ve had tremendous success supporting high-performance workloads in the energy sector. These results underscore the advantage of using Qumulo to accelerate business-critical subsurface workflows, helping customers reduce time-to-results and lower costs. We thank Cegal for partnering with us to showcase what’s possible.”
In single-workstation Petrel tests, CNQ completed workloads about 30 percent faster than Azure Files V2, a commonly used option for these scenarios. In multi-workstation tests, the gap widened significantly, with CNQ running 57 percent faster, highlighting its superior scalability in high-throughput, multi-client environments. These conditions mirror the demands of leading oil and gas exploration firms, where seismic datasets and geological surveys must be accessed simultaneously by engineers, geologists, and data scientists across global teams. All benchmarks were repeated multiple times, and CNQ consistently delivered predictable, reliable performance across every run.
While benchmarking, these workloads used less than two percent of available performance across IOPS and Throughput. This benchmark performance accelerated to the point where the results became client-side CPU-bound, demonstrating CNQ substantially improves application performance as customers scale the number of workstations accessing the storage. This means that customers will see even better performance if they have multiple, high-performance Virtual Machines (VMs) for their Petrel configurations. These tests were conducted on a 3-node L8sv3 cluster (base configuration) using Cloud Native Qumulo on Azure and a general-purpose hot configuration of Azure Files V2 – both systems considered “out of the box” deployments.
Both configurations were priced at approximately $52/TB/month for 500TB of usable capacity (pre-discount), notably lower than other file system alternatives in Azure, which often range from $75–$105/TB/month even after discounts. At a multi-petabyte scale, customers can achieve the same performance for as little as $30/TB/month using Qumulo, because of Qumulo’s unique architecture and available incentives.
With its performance, scalability, and cost efficiency, CNQ on Azure is an ideal platform for energy companies and geoscience teams running Petrel and other data-intensive workflows, especially as concurrency and data volumes increase.
