Everpure has introduced new storage solutions designed to help companies move AI pilot projects into production more quickly.
Everpure has introduced FlashBlade//EXA and the beta version of Everpure Data Stream planned for later in 2026.
The solutions aim to reduce the costs and complexity that currently prevent many AI projects from moving from pilot to production use. Evergreen//One for AI now includes FlashBlade//EXA to provide greater scalability and throughput for training and inference tasks. The Data Stream beta is designed to automate data pipelines from capture to inference and to eliminate manual intermediate steps.
“Many companies still view AI as just another workload. We support them in breaking down data silos and moving projects into production,” said Kaycee Lai, Vice President AI at Everpure.
Customers report practical benefits. Andrea Moccia, VP AI and Data at Options Technology, said Evergreen//One has simplified capacity planning. Storage can now be deployed worldwide and used on a pay-as-you-go basis. Sabur Mian, CEO of STN, emphasized performance consistency. With FlashBlade//EXA the company scaled from four to 192 nodes without significant performance drops.
Technically, FlashBlade//EXA aligns with NVIDIA’s modular STX reference architecture and is prepared for the Vera Rubin platform. It works with BlueField-capable controllers and context storage architectures. Benchmarks from SPECstorage Solution 2020 and MLPerf indicate that the system achieves GPU utilization above 90 percent in large NVIDIA Hopper clusters.
Everpure Data Stream was developed in collaboration with Supermicro and is based on the NVIDIA AI Data Platform reference design. The solution is intended to simplify data preparation and provisioning and supports platforms such as the NVIDIA RTX PRO 6000 Blackwell Server Edition. Support for the RTX PRO 4500 Blackwell Server Edition GPU is planned.
Everpure presents the combination of storage hardware and automated data orchestration as a foundation for continuous AI optimization in enterprises.
Dr. Jakob Jung is Editor-in-Chief of Security Storage and Channel Germany. He has been working in IT journalism for more than 20 years. His career includes Computer Reseller News, Heise Resale, Informationweek, Techtarget (storage and data center) and ChannelBiz. He also freelances for numerous IT publications, including Computerwoche, Channelpartner, IT-Business, Storage-Insider and ZDnet. His main topics are channel, storage, security, data center, ERP and CRM.
Contact via Mail: jakob.jung@security-storage-und-channel-germany.de