Hammerspace Unveils the Data System for the AI Era

Hammerspace today announced a major expansion of its technology portfolio with the Hammerspace Data System, extending its proven data unification, orchestration and acceleration capabilities with integrated data intelligence and high-performance storage. The result is one complete, integrated Data System for AI and HPC that spans existing and new infrastructure across data centers and clouds, delivering greater performance efficiency, simpler operations and more value from existing storage investments.

This press release features multimedia. View the full release here: https://www.businesswire.com/news/home/20261117928920/en/

The Hammerspace Data System: Five capabilities, one unified data system.

The Hammerspace Data System: Five capabilities, one unified data system.

Hammerspace is bringing to data the same simplicity that AI factories brought to compute. As open standards transform AI models and compute infrastructure, Hammerspace extends that interoperability to data, breaking down storage silos and enabling organizations to put their existing data to work for AI wherever it runs, without disrupting the applications and workflows that already depend on that data.

AI demands extreme performance, but achieving it by overprovisioning SSD capacity or adding in-band storage servers drives unnecessary cost, power consumption and performance bottlenecks. Hammerspace takes a fundamentally different approach: separating metadata from the data path, enabling high-performance parallel I/O, and leveraging GPU server-local NVMe for Tier 0 acceleration. Combined with intelligent data orchestration across existing infrastructure, clouds and neoclouds, this architecture delivers the performance AI demands with greater efficiency per dollar, watt and rack unit.

The expanded Hammerspace Data System builds on this foundation with three major advancements, adding integrated data intelligence, a new high-performance flash architecture and simplified delivery as a complete system.

  1. Data Intelligence: Discovers, catalogs, queries and acts on data across distributed infrastructure, turning data understanding into automated orchestration that delivers the right data to AI workloads.

  2. Data Storage: Hammerspace expands its data storage capabilities with the introduction of the Hammerspace Open Flash Platform (OFP), a DPU-based flash storage platform with the best performance density of any solution on the market. Hammerspace OFP adds a new high-performance storage option alongside existing third-party flash and cost-efficient HDD storage, all unified within a single data system. KV cache storage uses Tier 0 to unify server-local NVMe across a cluster into shared, high-performance context storage, enabling inference software to reuse previously computed context, reduce redundant computation and keep GPUs productive.

  3. Delivered as a System: Simplifies the purchase, deployment and operation of AI data infrastructure with a complete, integrated system combining Hammerspace software, metadata nodes and high-performance storage, while retaining the flexibility to deploy software-only or leverage existing infrastructure.

Hammerspace OFP: Maximum Performance Density. Minimum Infrastructure.

Hammerspace OFP shatters the traditional economics of high-performance AI storage through a fundamentally different architecture. By eliminating conventional x86 storage servers, separating metadata from the data path and connecting high-density flash directly to standard networks through specialized DPUs, OFP removes traditional performance bottlenecks and unlocks dramatically greater performance per rack unit.

OFP has demonstrated industry-leading performance density:

  • 390 GB/sec write throughput per rack unit – 4x that of the nearest competitor

  • 365 GB/sec read throughput per rack unit – 2x that of the nearest competitor

  • 15M IOPs per rack unit – more than 2x that of the nearest competitor

More performance in every rack unit. Less infrastructure to power it. This architectural efficiency reduces the need to overprovision expensive flash capacity simply to achieve throughput targets, while lowering hardware costs, power consumption and rack space. Integrated into the Hammerspace Data System, OFP delivers extreme performance efficiency for AI training, inference and context storage — without creating another storage silo.

“The race for AI performance isn’t slowing down. But the next era of AI infrastructure will be defined by how efficiently we deliver that performance — maximizing throughput while minimizing power, footprint and cost,” said David Flynn, Founder and CEO of Hammerspace. “Simply adding more storage isn’t the answer. AI factories simplified compute. We’re bringing that same transformation to data systems by unifying intelligence, orchestration, high-performance access and storage in one system. With Hammerspace OFP, we’re delivering more performance in less rack space and reducing storage infrastructure overhead, while giving organizations the freedom to put their data to work wherever AI runs.”

Five Capabilities. One Unified Data System.

The Hammerspace Data System integrates five capabilities into one architecture. Building on its proven strengths in data unification, data orchestration and data acceleration, Hammerspace adds data intelligence and data storage to deliver a complete system for AI and HPC.

  • Data Unification — Put existing data to work for AI faster. Unifies data across storage systems, sites and clouds in minutes, without disruptive migrations. Existing applications and AI workloads access the same data through standard protocols.

  • Data Intelligence — Find and activate more data. Catalogs and queries data across distributed infrastructure, connecting data intelligence directly to automated orchestration policies.

  • Data Orchestration — Deliver data wherever AI runs. Automates data placement across storage systems, data centers and clouds, eliminating manual staging and enabling workloads to run wherever compute is available.

  • Data Acceleration — Keep GPUs productive. Delivers high-performance parallel reads and writes, with Tier 0 using server-local NVMe as a shared data accelerator to eliminate bottlenecks and improve GPU utilization.

  • Data Storage — Maximize performance and capacity efficiency. Integrates Hammerspace OFP high-performance flash with existing and new third-party flash and HDD storage, scaling performance and capacity without creating new silos.

Now Available as a Complete, Integrated System

For the first time, customers can purchase the full Hammerspace Data System as one integrated solution, combining all five capabilities with Hammerspace software, metadata nodes and high-performance storage — installed and supported by Hammerspace. This simplifies purchasing, deployment and operations while delivering the flexibility of an open, standards-based architecture.

Customers can also deploy Hammerspace software as a software-defined offering on third-party hardware or leading cloud platforms, extending their existing infrastructure without creating new storage silos.

The Hammerspace OFP storage system is available to quote now and will ship in the first half of 2027.

To experience the Hammerspace Data System, visit booth #4238 at SC26 or contact Hammerspace to schedule a demonstration.

Supporting Quotes

“Hammerspace has been crucial for us to move toward one single data system. No matter where we want to do our training, it can be started from the same place, and we know the data is going to be there,” said Daan Rademaker, Staff Data and AI Platform Engineer, kaiko.ai.

“As organizations build AI infrastructure at massive scale, they need storage systems that deliver exceptional performance without the traditional tradeoffs in cost, power and footprint. A DPU architecture that puts every core in the data plane and runs standard Linux changes what’s possible, enabling a new level of performance and efficiency for AI storage. With the Xsight Labs E1 DPU inside Hammerspace OFP, we’re helping organizations keep GPUs productive, accelerate AI training and inference and scale more efficiently. We’re proud to partner with Hammerspace to bring this architecture to market,” said Eric Vallone, VP of Business Development and Head of Commercial, Xsight Labs.

“AI infrastructure is moving from fragmented technologies to integrated systems. Compute has made that leap. Data hasn’t. The next era of AI requires a unified system that brings intelligence, orchestration, high-performance access and storage together. Hammerspace is driving that shift, enabling organizations to unlock their existing data, eliminate infrastructure boundaries and deliver the performance AI demands — without creating another silo. This is how data infrastructure must evolve to power AI at scale,” said Simon Robinson, Chief Analyst for Storage and Data Infrastructure, Omdia

Resources

About Hammerspace

Hammerspace is an AI systems company that delivers a unified data management and storage system for AI. Combining high-performance storage, data management, and data intelligence, the Hammerspace Data System spans the entire data estate, including existing storage infrastructure, new storage, data centers, and clouds. Hammerspace helps organizations put all their data to work for AI, accelerate AI training and inference, and maximize the value of existing infrastructure. By combining high-performance Tier 0, flash, and HDD storage in one system, Hammerspace increases productivity, reduces infrastructure and operating costs, and gives organizations the flexibility to scale AI without creating more storage silos.

Learn more at www.hammerspace.com.

Hammerspace and the Hammerspace logo are trademarks of Hammerspace, Inc. All other trademarks used herein are the property of their respective owners.

©2026 Hammerspace, Inc. All rights reserved.

Media gallery