NetApp to acquire UK's PEAK:AIO for AI-scale metadata storage
NetApp (NASDAQ: NTAP) has announced its intent to acquire PEAK:AIO, a Manchester-based software-defined storage company whose technology is designed for high-performance, parallel AI workloads. The deal, which remains subject to customary regulatory approvals and closing conditions, is intended to accelerate NetApp's AI infrastructure roadmap by adding specialised metadata services and parallel namespace capabilities to its ONTAP platform.
PEAK:AIO's core innovation centres on disaggregating metadata from data storage, allowing metadata services to scale independently of underlying data volumes. The company says its platform supports environments spanning a single server up to exabyte-scale deployments on commodity hardware, using parallel NFS-based access. Current deployments include Los Alamos National Laboratory, NHS AIDE, Oxford Robotics Institute, Carnegie Mellon University and several UK universities.
The deal
NetApp chief executive George Kurian said the acquisition addresses a structural gap in enterprise storage: platforms designed for traditional workloads struggle to keep pace with the concurrency and throughput demands of GPU clusters running large AI models. The combined architecture, as the company describes it, would support trillions of files and multi-exabyte environments while preserving ONTAP's existing resilience and security properties.
Mark Klarzynski, founder and chief technology officer of PEAK:AIO, said the deal offers the startup access to NetApp's global customer base and operational maturity. No financial terms were disclosed. The absence of deal value or an acquisition price is notable given that PEAK:AIO's research-institution deployments suggest it remains a relatively early-stage commercial business rather than a scaled revenue generator.
NetApp has not set a timetable for integrating PEAK:AIO's technology into generally available ONTAP releases, noting in a statement of product direction that development timelines remain at its sole discretion.
Market context
The AI infrastructure storage market is attracting significant attention from a broad field of competitors. Pure Storage, IBM Storage Scale (formerly Spectrum Scale), Weka and VAST Data are all positioned in high-performance parallel file storage, and hyperscalers including AWS and Microsoft Azure offer managed file services optimised for GPU workloads. The competitive dynamic is shifting toward vendors that can credibly serve both the metadata performance layer and the underlying data fabric in a single integrated stack, rather than forcing customers to bolt together separate systems.
NetApp's move to acquire rather than build its metadata capability suggests the engineering challenge of scaling ONTAP's namespace architecture for AI-scale file counts is non-trivial. Parallel file systems with disaggregated metadata are a well-established concept in high-performance computing, but their adoption in mainstream enterprise AI environments has lagged behind GPU procurement cycles.
Regulatory and standards read-across
The transaction involves a UK-headquartered target, so NetApp will need to consider whether the deal triggers a review under the UK National Security and Investment Act 2021, which covers acquisitions of companies working in advanced computing and data infrastructure. The release does not reference any such filing, and the threshold criteria depend on PEAK:AIO's revenue and the nature of its government-sector contracts, including its NHS deployment. Editors may wish to seek comment from NetApp on NSI Act status before publication.
For enterprise buyers, the acquisition raises an integration horizon question: ONTAP customers evaluating AI storage platforms now face uncertainty about whether PEAK:AIO's standalone product line will continue to be developed independently or will be folded into a future ONTAP release. NetApp's statement of product direction disclaimer is notably cautious on this point.