Longtime Elite partners deepen decade-long relationship with certified integration and joint whitepaper on high-performance storage for enterprise AI
NEW YORK, September 15, 2026 – Presidio, a leading provider of AI business outcome solutions, today announced the full integration of Everpure into PATH Lab, the validation environment within Presidio’s PATH portfolio. The announcement coincides with the launch of PATH and the release of a joint technical whitepaper, “High-Performance Storage as the Backbone of AI Infrastructure.”
PATH Stack gives organizations greater control over how AI runs, where it runs, and what it costs, delivering a consistent, secure, and scalable AI foundation with a common operating model across on-premises, colocation, neo-cloud, and hyperscaler environments. PATH Lab is Presidio’s purpose-built environment for validating AI use cases against real enterprise requirements, closing the gap where most enterprise AI initiatives stall: the transition from proof of concept to production.
With this integration, clients can now test and validate AI workloads in PATH Lab on Everpure’s FlashBlade architecture, a disaggregated, all-NVMe, scale-out platform purpose-engineered for unstructured data and modern AI and machine learning workloads. Workloads are proven on the same storage foundation they will run on in production, with certified NVIDIA GPUDirect Storage support, unified file and object access, and throughput that increases as blades are added.
“Storage performance is the most overlooked variable in enterprise AI economics,” said Chris Cagnazzi, Chief Innovation Officer at Presidio. “GPUs waiting on data are capital sitting idle. Integrating Everpure into PATH Lab means our clients validate use cases on infrastructure that reflects production reality, with full visibility into performance and cost before they commit.”
“Every organization has an AI ambition, but very few have a foundation that can support it,” said Nirav Sheth, VP of Worldwide Systems Engineering at Everpure. “PATH flips that equation. Presidio in collaboration with Everpure has built an environment that brings real and measurable impact to private AI, making it a working architecture, instead of just a slide. Customers can keep their data, models and governance under their control from ingestion through inference, while also addressing the security, scalability and cost challenges that come with putting AI into production.”
The joint whitepaper, available today, examines why storage performance is foundational to AI success. Key findings include:
- Presidio infrastructure assessments typically find enterprise training environments running at 30 to 50 percent GPU utilization, with storage and pipeline performance one contributor among several
- Where storage is measured to be the binding constraint, remediation has produced 1.5x to 3x faster model training cycles and a 10 to 25 percent reduction in cost per model
- In those same engagements, GPU utilization moved into a 60 to 80 percent band and inference latency fell 15 to 40 percent
The whitepaper also provides a technical deep dive into the Everpure FlashBlade architecture, a reference architecture for an AI-optimized storage stack, and best practices for AI storage deployment.
Presidio has partnered with Everpure for more than 10 years and holds Elite partner status. The relationship spans joint engineering, certified reference architectures, and thousands of successful client engagements.
Click here to download the whitepaper.
To learn more about how we AI, visit presidio.com/how-we-ai
About Presidio
Presidio is a leading provider of AI business outcome solutions, helping organizations modernize IT, optimize performance, and accelerate innovation from edge to core to cloud and across modern networks. With full stack accountability from infrastructure to application, Presidio brings decades of enterprise trust to every engagement. Learn more at www.presidio.com.