cd /news/ai-infrastructure/ddn-exascaler-7-adds-a-software-defi… · home › topics › ai-infrastructure › article
[ARTICLE · art-172022] src=storagereview.com ↗ pub= topic=ai-infrastructure verified=true sentiment=↑ positive

DDN EXAScaler 7 Adds a Software-Defined Option, Starting on the AI2200 With Third-Party Appliances to Follow

DDN will sell its Lustre-based EXAScaler parallel file system as standalone software for the first time with the upcoming EXAScaler 7 release, starting on DDN's AI2200 server with qualified third-party appliances to follow, according to DDN Chief Product Officer Omer Asad. A limited beta is planned for early 2027, and DDN is separately adding multi-tenancy REST APIs with per-tenant quality-of-service limits, File Restore, an audit log, and intelligent hybrid I/O in the first half of 2027, though DDN has published no performance figures, named third-party hardware vendors, or detailed licensing. DDN also reported that RAVEL recorded 98.8 percent GPU utilization in a 24-hour training test on the first DDN Enterprise AI HyperPOD at the AMPHIX Center of Excellence in California, a result run on DDN Infinia rather than EXAScaler.

key takeaway

EXAScaler 7 will let teams run DDN's Lustre-based parallel file system on qualified third-party servers starting with the AI2200, with a limited beta in early 2027 and multi-tenancy REST APIs, File Restore, audit logging and hybrid I/O arriving in the first half of 2027.

by read4 min views3 publishedNov 18, 2026
DDN EXAScaler 7 Adds a Software-Defined Option, Starting on the AI2200 With Third-Party Appliances to Follow
Image: Storagereview (auto-discovered)

DDN will offer EXAScaler as software that customers install on qualified industry-standard servers, a change that arrives with the upcoming EXAScaler 7 release. The Lustre-based parallel file system has been sold only inside DDN’s purpose-built appliances or through cloud partners, and the new option starts on the DDN AI2200, with qualified third-party appliances to follow. A limited beta is planned for early 2027, and DDN is separately adding multi-tenancy, data protection, and performance features to EXAScaler in the first half of 2027.

Software-Defined Lustre for AI and HPC #

DDN describes EXAScaler as open Lustre with AI I/O optimization, automated deployment, and enterprise support. A parallel file system reads from many storage servers at once, which is how it keeps up with the small files, metadata-heavy reads, and frequent checkpoints that AI and HPC jobs generate. EXAScaler is the file system inside appliances such as the AI400X3M that DDN launched at ISC 2026, and DDN says it will keep building purpose-built systems for customers who want a complete DDN solution.

With EXAScaler 7, the same software runs on what DDN calls verified hardware from a growing ecosystem of suppliers, beginning with its AI2200 server. DDN pitches the option at hyperscalers and cloud providers that already have storage infrastructure, and at AI platform teams that want to pick hardware to suit their requirements. DDN has not named third-party hardware vendors, published minimum hardware requirements, or said how the software will be licensed.

Omer Asad, Chief Product Officer at DDN, tied the move to the rest of the lineup: “By making EXAScaler available as software, we’re extending that flexibility across our portfolio.” DDN Infinia, which gained multi-tenancy and POSIX support in version 2.4, is already available in a software-defined format, so both DDN data platforms will have that option. EXAScaler also remains available through cloud partners such as Google Cloud Managed Lustre.

Tenant Controls and Data Protection #

EXAScaler 7 will add multi-tenancy REST APIs that apply per-tenant quality-of-service limits and fair sharing, so several teams can share one system with less interference between their jobs. File Restore recovers accidentally deleted files, and an audit log supports investigations. On the performance side, DDN lists intelligent hybrid I/O for higher throughput and IOPS. The company has published no performance figures for these features and says validated data will follow.

Shared Data for Simulation and AI Training #

DDN’s example for the software-defined option is a site that runs HPC simulation and AI training against the same data without copying it between environments. In life sciences, researchers train AI surrogate models on the outputs of molecular dynamics simulations to predict protein behavior without rerunning every simulation. DDN has not given a general availability date for the software-defined deployment beyond the early 2027 beta, and it says all of the dates are targets that may change.

AMPHIX HyperPOD Test Reports 98.8 Percent GPU Utilization on Infinia #

DDN also published results from the first DDN Enterprise AI HyperPOD installed at AMPHIX, the RAVEL and Strata Expanse deployment model that named RT-One as its first commercial customer. HyperPOD is a packaged system built on Supermicro servers with NVIDIA accelerated computing and DDN Infinia as its storage, so the result applies to DDN’s other data platform and was not run on EXAScaler. In a 24-hour training test at the AMPHIX Center of Excellence in California, RAVEL recorded 98.8 percent GPU utilization with zero thermal throttling or GPU errors and 37C of thermal headroom at peak load.

DDN told us the system had eight NVIDIA RTX PRO 6000 GPUs and 500TB of Infinia running on Supermicro hardware. The job trained a large language model based on the GPT-2 architecture from 70TB of raw training data, with all of it read from and written back to storage over S3 across the network. GPUs held 95 percent utilization or higher for 98 percent of the run, and hardware telemetry was sampled once per second. DDN describes this as an AMPHIX internal test, and we did not run or audit the results.

Eight GPUs typically fit in a single server, so the test shows a small HyperPOD keeping its GPUs busy for a day while streaming training data over S3, and it leaves behavior at cluster scale unmeasured. DDN has not published the training throughput, the inference results RAVEL also collected, or the server and network configuration. AMPHIX is offering the system for customer proofs of concept at the California site, and Strata Expanse and RAVEL describe a plan to connect more than 100 sites.

── more in #ai-infrastructure 4 stories · sorted by recency
── more on @ddn 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/ddn-exascaler-7-adds…] indexed:0 read:4min 2026-11-18 · —