Pure Storage Unveils FlashBlade//EXA for AI and HPC

Pure Storage has introduced FlashBlade//EXA, a data storage platform designed for artificial intelligence (AI) and high-performance computing (HPC) workloads. Reportedly, The platform aims to address challenges associated with large-scale metadata operations, which are critical for high concurrency and intensive data processing in AI and HPC environments.

Addressing Storage Constraints in AI and HPC

The increasing complexity and scale of AI model training have led to evolving storage requirements. Traditional storage architectures, originally developed for high-performance computing, face limitations when applied to AI workloads, particularly in areas such as metadata performance, concurrent processing, and low-latency operations.

FlashBlade//EXA is reportedly designed to overcome these constraints by decoupling metadata and data scaling, allowing for high throughput and improved efficiency in managing AI-driven workloads. Initial testing indicates that the platform is capable of over 10 terabytes per second of read performance within a single namespace, positioning it among the highest-performing storage solutions available.

Scalability and Architecture

The FlashBlade//EXA architecture is built to scale both data and metadata independently. It incorporates third-party data nodes, enabling a more flexible and scalable performance model suited to AI and HPC demands. The use of standard networking protocols is intended to streamline deployment, management, and scalability.

With AI models processing diverse data types—text, images, and video—at high concurrency, traditional storage systems face challenges in handling asynchronous checkpointing, latency, and metadata management. The parallel, disaggregated architecture of FlashBlade//EXA would support workloads, helping improve GPU utilization and training and inference processes.

Key Technical Capabilities

FlashBlade//EXA integrates metadata engines with the Purity operating system, aiming to improve storage efficiency and performance. Reportedly, Key features include:

• Performance Scaling: Supports high-speed AI pipelines through parallel processing and scalable metadata IOPS, with a projected 10+ terabytes per second in a single namespace.

• Metadata Optimization: Designed to manage large-scale AI datasets with high availability and resilience, reducing manual tuning requirements.

• Infrastructure Integration: Supports industry-standard protocols and includes NVIDIA ConnectX NICs, Spectrum switches, and accelerated communication libraries for optimized AI and HPC processing.

The platform builds on Pure Storage’s previous innovations, aiming to provide storage solutions for organizations engaged in AI model development and large-scale computing operations.

“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, CTO, 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, VP, 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.

Read More:

Partners Lead the Way in Shaping the Future of Virtualization

Focus on Hybrid Cloud Integration and AI powered IT Automation