What happened
DDN introduced a software‑defined deployment model for its EXAScaler data platform, beginning with the upcoming EXAScaler 7 release. The new option lets enterprises install DDN’s high‑performance storage software on qualified third‑party servers—starting with the DDN AI2200—rather than requiring DDN‑built hardware or cloud‑partner solutions. A limited beta is planned for early 2027, and DDN will continue offering its traditional purpose‑built systems alongside the new model. The company also noted ongoing work on multi‑tenancy, performance, and data‑protection enhancements slated for general availability in the first half of 2027.
DDN announced that its EXAScaler platform, traditionally delivered as purpose‑built appliances or via cloud partners, will be offered as software that can run on industry‑standard servers. The initial hardware target is the DDN AI2200, with plans to certify additional third‑party appliances later.
The software‑defined deployment is tied to the upcoming EXAScaler 7 release. DDN will run a limited beta in early 2027 to validate the approach and gather performance data before broader availability.
The company emphasized that this option is intended for enterprises that need flexibility to choose hardware that aligns with specific AI or HPC workloads, regulatory constraints, or cost considerations. DDN will still provide its integrated hardware solutions for customers who prioritize turnkey performance and reliability.
DDN also highlighted ongoing development of EXAScaler enhancements—multi‑tenancy, performance tuning, and data‑protection features—targeted for general availability in the first half of 2027. These improvements aim to support mixed AI‑HPC environments and larger, shared‑resource deployments.
Why it matters
The move expands flexibility for AI and HPC customers who need to integrate storage with diverse architectures. By decoupling software from proprietary hardware, organizations can select servers that match specific performance, cost, or regulatory requirements, potentially reducing capital expense and simplifying upgrades. This is especially relevant as enterprises blend AI training, , and traditional HPC simulations, where moving large datasets between separate storage systems can be a bottleneck. DDN’s shift mirrors a broader industry trend toward software‑defined infrastructure, giving hyperscalers, national labs, and sovereign AI initiatives more options for building scalable AI factories. However, the offering remains limited to verified hardware and is still in beta, so performance parity with DDN’s purpose‑built systems is not yet confirmed.
Software‑defined storage allows organizations to avoid vendor lock‑in on hardware, potentially lowering total cost of ownership and enabling faster hardware refresh cycles.
For AI workloads that generate petabytes of training data, keeping data on a single platform that can serve both simulation outputs and AI training reduces data movement overhead and latency.
The announcement aligns with a broader industry shift toward modular, cloud‑like infrastructure in on‑premise data centers, supporting the growing need for hybrid AI deployments that span on‑prem, edge, and cloud environments.
DDN’s claim of “unmatched performance” will need independent verification in the software‑defined context; early benchmarks will be crucial to assess whether the flexibility trade‑off impacts throughput or latency compared with DDN’s proprietary systems.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
crm_get_transaction(id='4092').Which component of an AI application is the machine-learning model itself?
What to watch next
Key indicators to monitor include the performance benchmarks DDN releases for the software‑defined EXAScaler, adoption rates among early‑beta participants, and the timeline for expanding qualified third‑party hardware beyond the AI2200. Additionally, the rollout of the planned multi‑tenancy and data‑protection enhancements in early 2027 will affect how enterprises consolidate AI and HPC workloads on shared storage. Competitive responses from other storage vendors offering software‑defined options could also shape market dynamics.
Performance releases from DDN as the beta progresses, especially comparisons with existing purpose‑built EXAScaler appliances.
The list of certified third‑party hardware expanding beyond the AI2200, which will indicate how broadly the solution can be adopted.
Customer case studies or pilot deployments in sectors like life sciences, finance, or autonomous vehicles that illustrate real‑world benefits and any integration challenges.
Competitive moves from other storage vendors (e.g., NetApp, Dell EMC) offering similar software‑defined storage options for AI workloads.