MinIO Teams with Arrow for Hyperscale Storage Pods

MinIO today announced a new partnership with IT distributor Arrow Electronics to launch AIStor Pods, a new line of shrink-wrapped, validated, and modular hyperscale storage solutions. The first four-node, 1.1 PiB AIStor Pods will bear the Supermicro logo, but more OEMs are expected to join the program.

The AI revolution has turbocharged demand for storage, both on a macro and an individual level. Organizations that once might have needed for a petabyte of storage are now looking for tens or hundreds of petabytes to stash images, video, text and other forms of unstructured data that can be used for AI model training and inference. Some organizations are now creeing into the exascale range, which until fairly recently was a domain only a handful of hyperscalers tread.

As a developer of one of the most popular open source S3-compatible object stores, MinIO has watched this demand surge. A year ago, MinIO launched a reference architecture for creating DataPods, which gave OEMs a repeatable pattern for building massive MinIO storage arrays atop industry standard servers. Supermicro, Dell, and HPE were the initial partners.

Now MinIO is following up the DataPods reference architecture with its new AIStor Pods line. Whereas the DataPods architecture provided a blueprint for hyperscale storage, AIStor Pods are actual products that customers can order, and Arrow Electronics, the exclusive distributor for AIStor Pods, will work with Supermicro to deliver them.

The first AIStor pods solutions will be delivered on Hyper1U X86 servers from Supermicro. The partners are eyeing a starter pod that includes four nodes and delivers 1.1 PiB of usable capacity, and customers can scale up from there. Customers will be able to obtain the AIStor Pods via three-year subscriptions. MinIO will provide tier-one support through its MinIO SUBNET program, and engage hardware providers as needed.

It’s all about eliminating the challenges to obtaining the storage customers need, said Mahesh Patel, MinIO’s chief business officer.

“With AIStor pods, we have collapsed the process of procuring hardware and software and configuring it into one simple step, via a per-node, hardened SKU,” Patel tells BigDATAwire. “You size your capacity, select your hardware, and issue a single PO through Arrow. Arrow handles the integration and delivery. You rack it, connect it, and you’re live almost instantaneously. And when you need more, you just add nodes. It’s a straightforward path that makes moving fast on AI infrastructure much easier.”

The AIStor Pods are based on AIStor, the reimagined object storage solution that MinIO launched last year and which added AI-specific commands to the object storage software, such as the promptObject command, which allows customesr to “talk” to their data. The company is still engaging customers with its DataPods offering, and sees AIStor Pods as a complementary program.

Supermicro servers are a great fit for AIStor Pods, said Michael McNerney, Supermicro’s senior vice president of marketing and network security. “Customers want easy-to-deploy solutions and the pre-integrated and pre-validated AIStor on Supermicro servers provides that simplified deployment experience and can accelerate- time-to-value for customers,” he stated.

More hardware providers will be added to the AIStor Pods program in the future, Paptel said. The company could add more reference architures in the future. While Arrow is the sole distributor for AIStor Pods, expansion is expected as demand grows, he said.

As with DataPods architecture, the elmiantion of hardware lock-in is a key feature of AIStor Pods, Paptel said.

“We continuously evaluate new servers, CPU/NIC, and media options; the current pods reflect the best ROI and widely available choices today,” Patel told us. “If a customer opts for different hardware, the pod specs double as a solid reference design to compare against – same software, same requirements, no lock-in.

MinIO Pivots to AI with Launch of AIStor

MinIO Debuts DataPod, a Reference Architecture for Exascale AI Storage

The post MinIO Teams with Arrow for Hyperscale Storage Pods appeared first on BigDATAwire.