DDN Infinia / EXAScaler
by DDNUnited States
DDN Infinia data platform and EXAScaler parallel file system for AI workloads and HPC. Details according to the manufacturer.
Price on requestWe obtain the quote for you.
Free and without obligation. OROBO brokers — you buy directly from the manufacturer or its dealer.
- Manufacturer
- DDN
- Country
- United States · Chatsworth, CA
- Category
- Networking & storage for AI
- Guide price
- on request
About the category
What it is
Large AI systems are only as fast as their data access. Specialised storage systems and data platforms feed training data to many GPUs in parallel, save checkpoints and keep documents ready for language models. This includes networking that sustains high data rates between storage and servers.
What companies use them for
- Supplying training data to GPU clusters.
- Saving checkpoints of large training runs.
- Keeping company documents centrally for search and language models.
- Archiving image and sensor data from production and research.
What to look for when buying
- Throughput and parallel access to match the number of GPUs.
- Scalability: how does the system grow with data and users?
- Supported protocols and networks, such as NFS, S3 or RDMA.
- Data protection, encryption, tenant separation and support.
Integration & effortEffort: high
Storage for AI is planned together with the GPU servers and the network and is usually installed by the manufacturer or a partner. On your side are the data model, access rights, backup and the connection to the training and inference platform.
Software that fits
- Ceph (opens in a new tab)Open sourceDistributed storage for block, file and object data on standard servers.
- Kubernetes (opens in a new tab)Open source · Apache-2.0Distributes applications and workloads across servers and keeps them running.
- Prometheus (opens in a new tab)Open source · Apache-2.0Collects metrics such as utilisation and throughput and raises alerts when problems occur.
- Grafana (opens in a new tab)Open source · AGPL-3.0Dashboards that visualise metrics from servers, storage and applications.
Licence details without guarantee — the project’s own licence is authoritative. We can set up and integrate it for you at a fixed price.
Simulation
How it runs in your operation
A simplified, typical sequence — not a recording of a real deployment.
How much data, how fast, for how many GPUs?
Design storage, network and servers together.
Load datasets, assign permissions, set up backup.
Track throughput and utilisation, spot bottlenecks early.
Brokering
How brokering works
- 01Request
Add the product to your request and tell us quantity, goal and timing. Free and without obligation.
- 02OROBO checks and obtains the quote
We check your need and intended use, screen buyer and seller against EU and US sanctions lists and obtain the quote from the manufacturer or its dealer. Your contact details only go to the seller with your consent.
- 03You buy directly from the manufacturer or its dealer
Contract, payment, delivery and warranty run directly between you and the seller. OROBO is neither seller nor importer and does not accept payments for goods.
Free for you: the seller alone sets the price. OROBO may receive a commission from the seller — on request we tell you whether and how much; it influences neither recommendations nor order. For now we broker for buyers in the EU, the EEA, Switzerland and the United Kingdom. Terms § 6b · Disclosure
Similar products
Product details and prices come from manufacturers’ public information (as of 3 Nov 2026, foreign currencies converted at the ECB rate). Prices are non-binding, rounded guide values excluding VAT, shipping and duties; only the seller’s quote is binding. Category texts are general information and do not replace a case-by-case assessment.