How can Lenovo integrate its to-be-acquired Infinidat?

Analysis. Lenovo and Infinidat have been studiously quiet about how Lenovo’s to-be-acquired Infinidat storage will co-exist with Lenovo’s existing OEM’d NetApp ONTAP storage arrays. The Infinidat and NetApp hardware and software architectures are quite different and there is little compatibility between the two.

NetApp provides its clustered and dual controller-based ONTAP storage systems to a market spanning from small and medium business (SMB) to the upper enterprise market. Infinidat sells its 3-controller-based InfiniBox storage into the upper and high-end enterprise segments and there is an overlap with NetApp in the upper enterprise area;

Our understanding is that Lenovo will want to provide a degree of commonality and unity between the two storage systems. This will help it upsell its NetApp customer base to the Infinidat systems and cross-sell its NetApp systems into the Infinidat customer base. The current need for AI data pipelines to be based on a single view of a customer’s entire data estate is also going to encourage providing some integration between the Infinidat and NetApp environments.

What integration possibilities are there that could bring them closer together and provide Lenovo’s NetApp-based and Infinidat-based customers with a unified storage offering?

The obvious integration arena is software, with, initially, the two array systems being able to share data, and then, later operate under a unified management umbrella, and then have a federated data environment. This would provide cross system visibility through unified namespace metadata/data fabric ideas and data feeds to data lakes and AI data pipelines.

As they both support block, file, and S3 object storage protocols, data movers would need to be set up for each of these.

We understand Lenovo will have an InfiniBox development centre at Infinidat’s Israel-based facility. This is an obvious location for Infinidat-NetApp software and hardware integration efforts. Lenovo will clearly have free reign to dive deep into Infinidat’s system SW and hardware and develop it but, as it OEM’s NetApp’s systems, it will have more limited access to NetApp’s ONTAP’s software, with code access depth depending on the OEM licence.

Let’s look at data moving, system management, and namespace/data fabric possibilities in turn to check out the possibilities. We’ll look at this from an on-premises viewpoint primarily.

Data moving and access

Direct protocol-based access (e.g., mounting NetApp NFS exports from Infinidat clients) is possible but limited to read-only scenarios. This doesn’t provide a unified management framework.

NetApp ONTAP supports importing LUNs from third-party arrays via the FlexArray feature, and that would provide NetApp access to Infinidat-held data. But this is for block-level storage aggregation, not file-level access

Migration suppliers such as Datadobi and others, such as Komprise, could set up and implement a migration plan to move data between the InfiniBox and NetApp storage systems. and this could include block, file and object data. Replication-based suppliers like Cirata and Cirrus Data could move block data between the two environments. Cirrus Data has a partnership with Infinidat and provides block-level data migration using tools like Cirrus Migrate On-Premises (CMO).

We understand Commvault integrates with both vendors for migration. Customers could use Commvault’s Online Data Mobility (ODM) to configure source and target arrays, enabling snapshot-based migrations with automation for backups, recovery, and data movement. This supports petabyte-scale transfers and hybrid cloud.

Where there are hybrid on-prem+public cloud environments, customers use cloud as a bridge, with, for instance, replicating NetApp data to S3-compatible cloud storage (e.g., AWS or StorageGRID) using SnapMirror Cloud. Infinidat can then access this data via its AWS integrations or MinIO Gateway for S3-to-NAS translation, making it visible in InfiniBox as object or file storage. This avoids direct vendor integration but requires cloud connectivity.

System management

Here there is no compatibility. Both NetApp and Infinidat provide GUI and CLI-based management interfaces. They both have RESTful API access and that provides a development prospect. There would need to be an extension built to the Infinidat InfiniVerse software to cover the NetApp system management area. We can’t see NetApp agreeing to Lenovo developing extensions to NetApp’s control plane software to cover Infinidat kit.

Lenovo would need to extend its TruScale system moniutoring services to cover the Infinidat arrays, and that ought to be relatively easy.

Namespace and data fabric

The aim here is to provide a single logical or virtual view of the two separate physical environments. DataCore’s SANsymphony software virtualizes storage across heterogeneous environments, including NetApp and Infinidat’s InfiniBox. This creates a software-defined layer where, for example, NetApp data can be pooled, mirrored, or tiered into Infinidat-managed volumes for visibility and access. It supports synchronous/asynchronous replication and high availability, making data appear local to Infinidat applications. DataCore is an Infinidat technology partner.

A second possibility is to use Hammerspace’s Global Data Environment (GDE) as a bridging or federating layer. Infinidat and Hammerspace share a strategic technology partnership focused on enhancing data management and orchestration. It enables Infinidat customers to layer Hammerspace’s GDE on top of InfiniBox for file-based workloads, multi-site data access, and AI-optimized performance.

The GDE provides a parallel NFS-based file system that orchestrates data across Infinidat’s InfiniBox (block and file storage), creating a single, location-agnostic namespace for users and applications, allowing data to be indexed, tiered, and accessed as if local, regardless of where it’s stored. Hammerspace claims this combination supports the “world’s fastest file system for enterprise AI,” with low-latency access for large language models (LLMs) and GPU computing.

NetApp lists Hammerspace as a partner, so Hammerspace is readily available as a third-part data bridging and federation supplier.

AI

Lenovo has its AI Factory product set which uses NetApp-sourced DM and DG Series arrays, the DG Series being all-flash arrays with the DM Series being hybrid flash-disk arrays.

Lenovo is not yet OEM’ing NetApp’s newly announced, disaggregated AFX arrays. Nor does it have an AI data pipeline software stack sitting above its own arrays. There is one possibly in waiting, as NetApp has announced its AIDE (AI Data Engine) to provide an AI data pipeline to have AFX-stored data made deliverable to, and usable by, AI Large language Models (LLMs) and agents. AIDE operates at the file and object level for AI workflows, not the block level, and is, perhaps, a potential Lenovo OEM opportunity.

Infinidat has its own AI-related software with a RAG workflow architecture which links Infinidat systems to third-party AI software. The company is less advanced in the AI data pipeline area than NetApp, and this could well be an existing focus of its software development people. If it is not we might well think it should be.

Backup and data protection

Infinidat has its InfiniGuard data protection backup target and cyber-resilience system. An obvious opportunity is to sell this into Lenovo’s NetApp storage base, using Cohesity, Rubrik, and Veeam data protection offerings, as they support both NetApp and Infinidat.

InfiniGuard is a Veeam-ready repository. Cohesity can use InfiniGuard as a target tier for long-term retention or active access. Rubrik supports NetApp ONTAP for backups and is validated with Infinidat for petabyte-scale backups.

Selling InfiniGuard into Lenovo’s NetApp storage base looks like a relative no-brainer as NetApp doesn’t have its own backup target devices. It would provide an initial and bounded Infinidat footprint which could then be extended as integration facilities between the two storage systems were developed by Lenovo.

Lenovo could brand the InfiniGuard systems as another D-something sub-brand in its ThinkSystem Storage line, alongside its existing DG, DM and DE Series sub-brands; DI Series for example. This would be, we might think, better than having another, high-level, Think-something brand, like ThinkAgile for its HCI products, such as ThinkInfinite or, to be ridiculous for a moment, ThinkIdat.

Comment

Without wishing to disrespect Lenovo’s software capabilities, we can reasonably say that Lenovo has focussed more on delivering server, networking and storage boxes (hardware) than on developing its own software capabilities. By acquiring Infinidat, it will now have its own storage operating system and services capability.

The main focus of the Infinidat SW team is obviously to develop Infinidat-focussed software. We suggest a new and secondary focus will be, or could be, to work to bring joint services and management to cover both the Infinidat and OEM’d NetApp arrays. The quickest way to do that, in the short-term, would be use third-party bridging and federating software, and Hammerspace’s GDE looks well-placed to fulfill that role, particularly as it brings along data estate-covering AI data pipeline facilities as well.