Pure Storage (NYSE: PSTG) has unveiled FlashBlade//EXA, a cutting-edge data storage platform designed to support the growing demands of artificial intelligence (AI) and high-performance computing (HPC). With a focus on speed, scalability, and efficiency, the new solution aims to overcome limitations of traditional storage systems and optimise performance for GPU-intensive workloads.
Breaking performance barriers in AI storage
As AI models become more sophisticated, they require faster and more efficient data storage solutions. Traditional storage architectures struggle to keep up with high-speed data access needs, leading to bottlenecks that slow down model training and inference. Rob Lee, Chief Technology Officer at Pure Storage, highlighted the significance of the new platform, stating, โFlashBlade//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.โ
Initial testing indicates that FlashBlade//EXA can achieve read speeds of over 10 terabytes per second in a single namespace, setting a new benchmark in AI storage performance. Its ability to scale data and metadata independently ensures smooth handling of complex workloads, while the use of off-the-shelf third-party data nodes allows for seamless expansion. By leveraging standard protocols and networking, the platform also simplifies deployment and management for enterprises.
Adapting to modern AI and HPC requirements
The rise of AI has transformed storage demands, requiring systems that can process vast amounts of multimodal dataโincluding text, images, and videosโat high speeds. Traditional storage solutions, originally designed for predictable HPC environments, struggle with the diversity and concurrency required by modern AI applications.
Rob Davis, Vice President of Storage Networking Technology at NVIDIA, stressed the importance of high-performance storage in AI development: โData is the fuel for enterprise AI factories, directly impacting performance and reliability of AI applications. With NVIDIA networking, the FlashBlade//EXA platform enables organisations 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.โ
FlashBlade//EXA is purpose-built to support these evolving needs. Its disaggregated, massively parallel architecture ensures continuous storage performance as AI models grow in complexity. By improving GPU utilisation and reducing idle time, enterprises can accelerate their AI and HPC projects more effectively. The integration of Pure Storageโs metadata engine with its Purity operating system, along with cost-effective data nodes, also delivers a strong balance between price and performance.
Key features and industry impact
FlashBlade//EXA is designed to offer three main advantages:
- Unmatched performance at scale โ Achieves over 10 terabytes per second read speeds within a single namespace, reducing AI model training times and improving workflow efficiency.
- Reduced operational complexity โ Eliminates metadata bottlenecks, enhancing data availability and system resilience without manual intervention.
- Accelerated AI and HPC innovation โ Supports high-speed networking through NVIDIA ConnectX NICs, Spectrum switches, LinkX cables, and optimised communication libraries for superior performance.
Pete Manca, President of Penguin Solutions, described how the new storage solution meets the needs of modern AI environments: โAt 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 todayโs large data centres. We see it as an ideal solution for intensive AI environments and HPC applications.โ
Industry analysts also recognise the impact of AI on storage demands. Matt Kimball, Vice President and Principal Analyst at Moor Insights & Strategy, explained: โAI 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 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 organisations of all sizes.โ
The new storage platform is expected to be available by mid-2025.
