Pure Storage Unveils FlashBlade//EXA: A Revolutionary Storage Platform for AI and HPC Workloads

Pure Storage targets 'AI Factory' market with FlashBlade//EXA

As companies continue to feed more and more data into AI models, storage has emerged as a serious constraint for some. According to Pure Storage, those in search of value are also realizing that conventional storage systems are struggling to meet the demands of these AI workloads - in a world where GPUs are increasingly being deployed. Recognizing this opportunity, and in an attempt to keep ahead of the market, Pure Storage has announced its FlashBlade//EXA platform, which focuses on what the company terms the "AI factory" segment - organizations operating thousands of GPUs with storage requirements that fall between typical enterprise and hyperscale environments. The growth in AI model size and complexity has necessitated a shift in storage requirements, according to the vendor. It adds that traditional storage systems, designed for more predictable workloads, face limitations when applied to large-scale AI and high-performance computing (HPC). Speaking with Patrick Smith, Field CTO for EMEA at Pure Storage, he explained: While AI is advancing quickly, storage is struggling to keep up. The challenge is underutilized GPUs. Organizations join queues to acquire GPUs, and when they finally get them, they can't maximize their use due to performance limitations of back-end storage systems. Pure Storage claims that this inefficiency presents a problem. With GPUs requiring substantial investments, having them inactive while waiting for data impacts resource efficiency and can slow development cycles. With the AI arms race in full swing, where companies and political powers have a sense that 'the first to value will be the long term winner', high performing storage could well be key. Smith told diginomica that there are three main areas where storage struggles with AI demands: In particular, Pure is building on its previous FlashBlade technology to tackle the metadata challenge in AI scenarios: We often find higher throughput requirements on metadata than on the underlying data itself. This is becoming a more pressing issue for users of large AI systems. Specifically, Pure Storage has identified a market segment between traditional enterprise AI deployments and hyperscaler environments, which it calls the "AI factory". Smith said that enterprise AI deployments typically involve fewer than 1,000 GPUs in a single environment, with storage needs between 50 terabytes and 100 petabytes, and throughput requirements between 100 gigabytes and one terabyte per second. As we've seen in recent months, these enterprise environments focus on retrieval-augmented generation (RAG), medium-scale inference, data lakes, and mid-range machine learning workloads. At the other end, hyperscalers operate tens of thousands of GPUs with storage needs in the tens to hundreds of exabytes and throughput requirements exceeding 50 terabytes per second. The AI factory market falls between these two groups, comprising of GPU clouds, high-end enterprises, tech companies, and AI-native organizations. AI factories, according to Pure, are organizations that typically operate thousands of GPUs, require storage capacities up to multiple exabytes, and need throughput between 1-50 terabytes per second (typically 10+ terabytes). Pure Storage's understanding of the AI factory segment has been informed by existing partnerships, including one with CoreWeave, a GPU-as-a-service provider. Smith said: That experience has led us to realize that we have this opportunity, and they have that demand for greater scale. In addition to companies like CoreWeaver, Pure has also had interest from public sector initiatives, such as the European Union's plans for AI supercomputers across member states. Pure Storage has been in discussions with providers bidding on these projects to understand their requirements for performance, scaling, and storage flexibility. Smith said: We realized that we didn't have a product aligned with this segment. FlashBlade//EXA is designed to expand us out of enterprise AI and become more relevant for the AI factory. FlashBlade//EXA's main architectural feature is its disaggregated design, which separates metadata management from data storage, allowing independent scaling of each. The platform consists of a metadata core built on Pure Storage's FlashBlade technology, optimized for metadata performance, separating metadata out from the underlying data. This builds on the company's existing metadata processing capabilities. Smith said: In FlashBlade we built a metadata engine with strong performance and scalability. With FlashBlade//EXA, we're separating how we store the metadata from how we store the underlying data. The compute cluster communicates with the metadata core over NFS with TCP using parallel NFS. For data storage, the platform uses dedicated data nodes - essentially Linux boxes with substantial storage and high-performance networking. The compute cluster communicates with these data nodes using NFSv3 over RDMA (Remote Direct Memory Access). Initially, customers can source these data nodes from their preferred hardware manufacturer, with Pure Storage providing configuration recommendations for CPUs, cores, network cards, and flash drives. However, Pure Storage plans to release its own data nodes later in 2025, taking advantage of its Direct Flash Module technology. According to Smith, preliminary performance testing indicates FlashBlade//EXA will deliver more than 10 terabytes per second read performance in a single namespace - which Pure Storage says is a new benchmark for storage performance. He added: As you scale out this solution, you'll effectively be building it out in almost in units of 10 terabytes per second. It allows for 20 times as many files in a single namespace as a typical parallel file system. Smith added that Pure Storage expects FlashBlade//EXA general availability in summer 2025, with one early enhancement to include support for S3 object storage over RDMA: We're seeing S3 as an important trend in this space. I've had many customer conversations this year where they're looking at the difference between high-performance file and high-performance S3 - and S3 is becoming a real interest to them. FlashBlade//EXA represents Pure Storage's move into a new and growing segment of the AI infrastructure market. As and when enterprises shift from experimental AI to production-scale deployments, and as specialized AI cloud providers expand, the demand for storage solutions that can efficiently support thousands of GPUs is likely to increase. Pure Storage's identification of the "AI factory" as a distinct segment with specific requirements reflects the evolving AI market beyond simple enterprise/hyperscaler categories. According to Pure, the ability to scale data and metadata independently, essentially allows organizations to expand only the components they need rather than overprovisioning less-utilized resources, potentially offering cost benefits for growing deployments. An attractive prospect for buyers that are concerned by both efficiency and cost across AI workloads. As diginomica has noted previously, Pure has also secured a design win with one of the 'big four' hyperscalers - which was this week revealed to be Meta. It's clear from that announcement and the FlashBlade//EXA announcement that Pure is targeting storage deals well beyond the traditional enterprise - likely as it recognizes that in an 'AI world', there will be high storage requirements amongst this top end of the market. Enterprises may be using AI extensively in the future, but the storage demands for AI factories and hyperscalers are likely to be far more intense and wide-reaching, which presents a significant revenue opportunity for Pure Storage. If it can embed itself across all market segments, it could be on to a winner.

With FlashBlade//EXA, Pure Storage helps enterprises keep data-hungry GPUs fed to accelerate AI workloads - SiliconANGLE

With FlashBlade//EXA, Pure Storage helps enterprises keep data-hungry GPUs fed to accelerate AI workloads Data center infrastructure provider Pure Storage Inc. says it's stepping up to take care of the industry's rising demand for high-performance storage systems that can scale to keep up with the most advanced artificial intelligence models. The company today announced a new and dedicated data storage software platform called FlashBlade//EXA, saying it's designed to solve the "metadata bottleneck" problem that holds back a growing number of enterprises from scaling up their AI workloads. Designed for high concurrency, FlashBlade//EXA is uniquely able to scale regular data and its associated metadata independently. Even better, it works with any kind of storage array, while deployment is simplified through the use of standard protocols and networking. The company is trying to fix a major bottleneck in data center storage today amid the realization that legacy storage systems are simply unable to process data for AI workloads efficiently. Storage needs to be able to keep up with the computational intensity and volume of the graphics processing units that process the data needed for AI, but existing systems have been left wanting. The issue is that they face major limitations in terms of parallel and concurrent reads and writes. Existing storage systems have been optimized for traditional application environments, where the workloads are more predictable, so the focus has been on scaling up raw performance. But the vast majority of new AI workloads are more complex and require multimodal data, including text, images and videos that must be processed simultaneously by thousands of GPUs all working together. As such, AI workloads require not just massive performance, but also a massively parallel architecture that can keep pace with the underlying metadata associated with those massive volumes of multimodal data. The problem stems from the fact that AI simply doesn't look anything like traditional enterprises workloads, which are what almost every storage array is designed to address today, said Steve McDowell of NAND Research Inc. "AI requires a massively parallel architecture with a global namespace that can serve data up to thousands of GPUs simultaneously," he explained. "You never want to starve an expensive GPU of data. It requires a fundamentally different approach." McDowell said it's a problem that's well-understood by every player in the storage industry, with the likes of Dell Technologies Inc. and NetApp Inc. announcing solutions that intend to address it. But with FlashBlade//EXA set to launch in the summer, it looks like Pure Storage has beaten them in the race to bring its product to the market. The company explained that FlashBlade//EXA is purpose-built to address AI workloads, offering unmatched performance alongside sophisticated metadata management capabilities. With its disaggregated, massively parallel architecture, it solves the problem of scaling up AI workloads, eliminating idle GPU time so enterprises can accelerate AI training and inference. What's most surprising is that Pure Storage has made FlashBlade//EXA compatible with third-party data nodes, which means customers won't have to buy the company's expensive, flash-based storage arrays to benefit from it. "This will make adoption easy for hyperscalers, specialty-cloud providers and big enterprises that already have a heavy investment in storage," McDowell said. "It's a smart play for Pure that blends the best of hardware and software-first approaches. It also teases a future where Pure is as equally focused on selling into the hyperscale world as traditional enterprise." Customers have not yet had a chance to verify FlashBlade//EXA's performance capabilities, but Pure Storage said preliminary tests by early adopters show it can deliver more than 10 terabytes per second of read performance in a single namespace, dramatically improving on what traditional systems are capable of. "With FlashBlade//EXA, Pure is building a system that separates out metadata handling from the data itself," McDowell said. "This looks like a system built for AI training and it's delivering some stellar initial performance numbers in the process."

Pure Storage Unveils FlashBlade//EXA to Revolutionize AI and HPC

A modern storage solution must provide a massively parallel, disaggregated architecture to deliver flexibility at scale - ensuring that storage helps accelerate the pace of AI. A Modern Storage Architecture for AI and HPC Workloads Legacy high-performance storage architectures were optimized for traditional HPC environments, with more predictable and regular workloads and a focus on raw performance scaling. Today's AI workloads are complex and multi-modal - including text, images, videos - being processed simultaneously by tens of thousands of GPUs. This dramatic shift demands advanced metadata optimization alongside massive performance scaling to efficiently manage diverse data types and high concurrency.

Pure Storage Introduces FlashBlade//EXA™, the World's Most Powerful Data Storage Platform for AI and High-Performance Computing

Purpose-built to bring unmatched performance, scalability, and power of FlashBlade technology to next-generation, GPU-intensive AI and HPC workloads Pure Storage® (NYSE: PSTG), the IT pioneer that delivers the world's most advanced data storage technology and services, today debuted FlashBlade//EXA™, the industry's highest performing data storage platform engineered for the most demanding requirements of AI and high-performance computing (HPC). While legacy approaches to data storage have held back AI potential, FlashBlade//EXA breaks the metadata bottleneck with a proven architecture based on FlashBlade, built for high concurrency and the massive amounts of metadata operations typical of large-scale AI and HPC workloads. In preliminary testing, FlashBlade//EXA is projected to deliver more than 10 terabytes per second read performance in a single namespace - setting a new bar as the industry's highest-performing storage solution. The FlashBlade//EXA architecture scales data and metadata independently; provides near unlimited scale with off the shelf, third-party data nodes that enable highly scalable multi-dimensional performance; and reduces complexity in deployment, management, and scaling through the use of standard protocols and networking. Driving a Paradigm Shift in Storage More powerful GPUs have increased the pace and scale that foundational AI models can be trained. The explosive growth in model size and sophistication is driving a paradigm shift in storage requirements, where solutions must seamlessly keep up with the computational intensity and volume, and variety in data demands of AI and HPC. Legacy storage systems were not designed to meet modern AI requirements. When applied to large-scale AI and HPC, they face critical limitations with parallel and concurrent reads and writes, metadata performance, ultra-low latency, asynchronous checkpointing, and predictable, high throughput. A modern storage solution must provide a massively parallel, disaggregated architecture to deliver flexibility at scale - ensuring that storage helps accelerate the pace of AI. A Modern Storage Architecture for AI and HPC Workloads Legacy high-performance storage architectures were optimized for traditional HPC environments, with more predictable and regular workloads and a focus on raw performance scaling. Today's AI workloads are complex and multi-modal - including text, images, videos - being processed simultaneously by tens of thousands of GPUs. This dramatic shift demands advanced metadata optimization alongside massive performance scaling to efficiently manage diverse data types and high concurrency. FlashBlade//EXA is purpose-built for the challenges of AI workloads with unmatched performance and metadata management. Its disaggregated, massively parallel architecture enables storage flexibility at scale. Enterprises can adapt to evolving multimodal models, optimize reliability, and eliminate idle time to accelerate AI model training and inference while improving GPU utilization. Additionally, the combination of Pure Storage's metadata engine and Purity operating system with cost-efficient, off-the-shelf data nodes enables enterprises to achieve an unparalleled price-to-performance ratio. FlashBlade//EXA is built on a decade of Pure Storage innovation to provide transformational storage for AI and HPC environments. FlashBlade//EXA will: Deliver industry-leading performance at scale, leveraging Pure Storage's proven metadata capabilities to maximize AI pipeline efficiency and minimize delays during training and inference workflows. Enabling multidimensional performance with massively parallel processing and scalable metadata IOPS to support high-speed AI requirements. Providing 10+ terabytes per second in a single namespace, delivering the industry's best performance. Reduce management complexity by eliminating metadata bottlenecks Providing high metadata performance, availability, and resiliency to handle massive AI datasets without manual tuning or additional configuration. Accelerate AI innovation with a highly configurable and disaggregated architecture Powering the evolving AI and HPC landscape with industry standard protocols. Incorporating high-speed NVIDIA ConnectX NICs, Spectrum switches, LinkX cables, and accelerated communications libraries. FlashBlade//EXA is expected to become available in mid-2025. Executive Insight "FlashBlade//EXA delivers a massively parallel architecture that enables independent scaling of data and metadata to provide customers with unmatched performance, scalability, and adaptability for some of the largest, most demanding data environments in the world. Storage is now accelerating the pace of large-scale HPC and AI evolution," said Rob Lee, Chief Technology Officer, Pure Storage. "Data is the fuel for enterprise AI factories, directly impacting performance and reliability of AI applications. With NVIDIA networking, the FlashBlade//EXA platform enables organizations to leverage the full potential of AI technologies while maintaining data security, scalability, and performance for model training, fine tuning, and the latest agentic AI and reasoning inference requirements," said Rob Davis, Vice President, Storage Networking Technology, NVIDIA. "At Penguin Solutions, we are experts at helping our customers deploy AI infrastructure at scale. We know the complexity involved, and that legacy storage systems can undermine high-performing AI infrastructure. FlashBlade//EXA is purpose-built for the demands of modern AI and HPC workloads. It provides the bandwidth, manageability at scale, and configurable, disaggregated architecture needed to power today's large data centers. We see it as an ideal solution for intensive AI environments and HPC applications," said Pete Manca, President, Penguin Solutions. "AI has disrupted the storage market. Legacy storage environments are unable to handle the massive parallelism required of AI and HPC. Modern storage and data platforms must continuously feed the GPUs and accelerators training the large foundation models and support the most demanding workloads. However, with high performance storage comes complexity. With FlashBlade//EXA, Pure Storage is leveraging its decade of experience unlocking the potential of metadata performance, while abstracting the complexity associated with managing these environments. If FlashBlade//EXA delivers on its promise, it will return real value to organizations of all sizes," said Matt Kimball, VP & Principal Analyst, Moor Insights & Strategy. Resources To learn more about FlashBlade//EXA, please visit our website, and read this blog. About Pure Storage Pure Storage (NYSE: PSTG) delivers the industry's most advanced data storage platform to store, manage, and protect the world's data at any scale. With Pure Storage, organizations have ultimate simplicity and flexibility, saving time, money, and energy. From AI to archive, Pure Storage delivers a cloud experience with one unified Storage as-a-Service platform across on premises, cloud, and hosted environments. Our platform is built on our Evergreen architecture that evolves with your business - always getting newer and better with zero planned downtime, guaranteed. Our customers are actively increasing their capacity and processing power while significantly reducing their carbon and energy footprint. It's easy to fall in love with Pure Storage, as evidenced by the highest Net Promoter Score in the industry. For more information, visit www.purestorage.com.

Pure Storage Introduces FlashBlade//EXA™, the World's Most Powerful Data Storage Platform for AI and High-Performance Computing

Purpose-built to bring unmatched performance, scalability and power of FlashBlade technology to next-generation, GPU-intensive AI and HPC workloads Pure Storage® (NYSE: PSTG), the IT pioneer that delivers the world's most advanced data storage technology and services, today debuted FlashBlade//EXA™, the industry's highest performing data storage platform engineered for the most demanding requirements of AI and high-performance computing (HPC). While legacy approaches to data storage have held back AI potential, FlashBlade//EXA breaks the metadata bottleneck with a proven architecture based on FlashBlade, built for high concurrency and the massive amounts of metadata operations typical of large scale AI and HPC workloads. In preliminary testing, FlashBlade//EXA is projected to deliver more than 10 terabytes per second read performance in a single namespace - setting a new bar as the industry's highest-performing storage solution. The FlashBlade//EXA architecture scales data and metadata independently; provides near unlimited scale with off the shelf, third-party data nodes that enable highly scalable multi-dimensional performance; and reduces complexity in deployment, management, and scaling through the use of standard protocols and networking. Driving a Paradigm Shift in Storage More powerful GPUs have increased the pace and scale that foundational AI models can be trained. The explosive growth in model size and sophistication is driving a paradigm shift in storage requirements, where solutions must seamlessly keep up with the computational intensity and volume, and variety in data demands of AI and HPC. Legacy storage systems were not designed to meet modern AI requirements. When applied to large-scale AI and HPC, they face critical limitations with parallel and concurrent reads and writes, metadata performance, ultra-low latency, asynchronous checkpointing, and predictable, high throughput. A modern storage solution must provide a massively parallel, disaggregated architecture to deliver flexibility at scale - ensuring that storage helps accelerate the pace of AI. A Modern Storage Architecture for AI and HPC Workloads Legacy high-performance storage architectures were optimized for traditional HPC environments, with more predictable and regular workloads and a focus on raw performance scaling. Today's AI workloads are complex and multi-modal - including text, images, videos - being processed simultaneously by tens of thousands of GPUs. This dramatic shift demands advanced metadata optimization alongside massive performance scaling to efficiently manage diverse data types and high concurrency. FlashBlade//EXA is purpose-built for the challenges of AI workloads with unmatched performance and metadata management. Its disaggregated, massively parallel architecture enables storage flexibility at scale. Enterprises can adapt to evolving multimodal models, optimize reliability, and eliminate idle time to accelerate AI model training and inference while improving GPU utilization. Additionally, the combination of Pure Storage's metadata engine and Purity operating system with cost-efficient, off-the-shelf data nodes enables enterprises to achieve an unparalleled price-to-performance ratio. FlashBlade//EXA is built on a decade of Pure Storage innovation to provide transformational storage for AI and HPC environments. FlashBlade//EXA will: Deliver industry-leading performance at scale, leveraging Pure Storage's proven metadata capabilities to maximize AI pipeline efficiency and minimize delays during training and inference workflows. Enabling multidimensional performance with massively parallel processing and scalable metadata IOPS to support high-speed AI requirements. Providing 10+ terabytes per second in a single namespace, delivering the industry's best performance. Reduce management complexity by eliminating metadata bottlenecks Providing high metadata performance, availability and resiliency to handle massive AI datasets without manual tuning or additional configuration. Accelerate AI innovation with a highly configurable and disaggregated architecture Powering the evolving AI and HPC landscape with industry standard protocols. Incorporating high-speed NVIDIA ConnectX NICs, Spectrum switches, LinkX cables, and accelerated communications libraries. FlashBlade//EXA is expected to become available in mid-2025. Executive Insight "FlashBlade//EXA delivers a massively parallel architecture that enables independent scaling of data and metadata to provide customers with unmatched performance, scalability, and adaptability for some of the largest, most demanding data environments in the world. Storage is now accelerating the pace of large-scale HPC and AI evolution," said Rob Lee, Chief Technology Officer, Pure Storage. "Data is the fuel for enterprise AI factories, directly impacting performance and reliability of AI applications. With NVIDIA networking, the FlashBlade//EXA platform enables organizations to leverage the full potential of AI technologies while maintaining data security, scalability, and performance for model training, fine tuning, and the latest agentic AI and reasoning inference requirements," said Rob Davis, Vice President, Storage Networking Technology, NVIDIA. "At Penguin Solutions, we are experts at helping our customers deploy AI infrastructure at scale. We know the complexity involved, and that legacy storage systems can undermine high-performing AI infrastructure. FlashBlade//EXA is purpose-built for the demands of modern AI and HPC workloads. It provides the bandwidth, manageability at scale, and configurable, disaggregated architecture needed to power today's large data centers. We see it as an ideal solution for intensive AI environments and HPC applications," said Pete Manca, President, Penguin Solutions. "AI has disrupted the storage market. Legacy storage environments are unable to handle the massive parallelism required of AI and HPC. Modern storage and data platforms must continuously feed the GPUs and accelerators training the large foundation models, and support the most demanding workloads. However, with high performance storage comes complexity. With FlashBlade//EXA, Pure Storage is leveraging its decade of experience unlocking the potential of metadata performance, while abstracting the complexity associated with managing these environments. If FlashBlade//EXA delivers on its promise, it will return real value to organizations of all sizes," said Matt Kimball, VP & Principal Analyst, Moor Insights & Strategy.