Pure Storage introduces data storage platform for AI and high-performance computing
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 optimised 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, optimise reliability, and eliminate idle time to accelerate AI model training and inference while improving GPU utilisation. 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 maximise AI pipeline efficiency and minimise 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.
leveraging Pure Storage's proven metadata capabilities to maximise AI pipeline efficiency and minimise delays during training and inference workflows.
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 summer 2025.
'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.
Resources
To learn more about FlashBlade//EXA, please visit our website, and read this blog.
Image Credit: Pure Storage
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Channel Post MEA
an hour ago
- Channel Post MEA
Fujitsu Starts Developing Quantum Computer With More Than 10,000 Qubits
Fujitsu has announced that it has started research and development towards a superconducting quantum computer with a capacity exceeding 10,000 qubits. Construction is slated for completion in fiscal 2030. The new superconducting quantum computer will operate with 250 logical qubits and will utilize Fujitsu's innovative 'STAR architecture,' an early-stage fault-tolerant quantum computing (early-FTQC) architecture also developed by the company. Fujitsu aims to make practical quantum computing possible, particularly in areas like materials science where complex simulations can unlock groundbreaking discoveries, and to this end will focus on advancing key scaling technologies across various technical domains. As part of this effort, Fujitsu has been selected as an implementing party for the 'Research and Development Project of the Enhanced Infrastructures for Post-5G Information and Communication Systems,' publicly solicited by the NEDO (New Energy and Industrial Technology Development Organization). Fujitsu will be contributing to the thematic area of advancing the development of quantum computers towards industrialization. The project will be promoted through joint research with Japan's National Institute of Advanced Industrial Science and Technology (AIST) and RIKEN, and will run until fiscal year 2027. Fujitsu is committed to driving forward the development of practical and industrialized quantum computing solutions. After this 10,000-qubit machine is built, Fujitsu will further pursue advanced research initiatives targeting the integration of superconducting and diamond spin-based qubits from fiscal 2030 and aims to realize a 1,000 logical qubit machine in fiscal 2035 while considering the possibility of multiple interconnected quantum bit-chips. Vivek Mahajan, Corporate Executive Officer, Corporate Vice President, CTO, in charge of System Platform, Fujitsu Limited, commented: 'Fujitsu is already recognized as a world leader in quantum computing across a broad spectrum, from software to hardware. This project, led by NEDO, will contribute significantly to Fujitsu's goal of further developing a Made-in-Japan fault tolerant superconducting quantum computer. We would also be aiming to combine superconducting quantum computing with diamond spin technology as part of our roadmap. By realizing 250 logical qubits in fiscal 2030 and 1,000 logical qubits in fiscal 2035, Fujitsu is committed to leading the path forward globally in the field of quantum computing. Additionally, Fujitsu will be developing the next generation of its HPC platform, using its FUJITSU-MONAKA processor line, which will also power FugakuNEXT. Fujitsu will further integrate its platforms for high-performance and quantum computing to offer a comprehensive computing platform to our customers.' Technology Development Focus Areas Fujitsu's research efforts will focus on developing the following scaling technologies: High-throughput, high-precision qubit manufacturing technology : Improvement of the manufacturing precision of Josephson Junctions, critical components of superconducting qubits which minimize frequency variations. : Improvement of the manufacturing precision of Josephson Junctions, critical components of superconducting qubits which minimize frequency variations. Chip-to-chip interconnect technology : Development of wiring and packaging technologies to enable the interconnection of multiple qubit chips, facilitating the creation of larger quantum processors. : Development of wiring and packaging technologies to enable the interconnection of multiple qubit chips, facilitating the creation of larger quantum processors. High-density packaging and low-cost qubit control : Addressing the challenges associated with cryogenic cooling and control systems, including the development of techniques to reduce component count and heat dissipation. : Addressing the challenges associated with cryogenic cooling and control systems, including the development of techniques to reduce component count and heat dissipation. Decoding technology for quantum error correction: Development of algorithms and system designs for decoding measurement data and correcting errors in quantum computations. Fujitsu's commitment to quantum computing is underscored by its ongoing R&D efforts. In August 2024, in collaboration with the University of Osaka, Fujitsu unveiled its STAR architecture, a highly efficient quantum computing architecture based on phase rotation gates. This architecture paves the way for early-FTQC systems capable of outperforming conventional computers with only 60,000 qubits. On the hardware front, the RIKEN RQC-Fujitsu Collaboration Center, established in 2021 with RIKEN, has already yielded a 64-qubit superconducting quantum computer in October 2023, followed by a world-leading 256-qubit system in April 2025.
Arabian Post
2 hours ago
- Arabian Post
SSSTC Launches World's First Industrial M.2 SSD Featuring KIOXIA's 8th generation BiCS FLASH™ Technology and PCIe® 5.0 Interface
The SSSTC CA8 Series is the world's first industrial M.2 SSD to combine KIOXIA's 8th-generation BiCS FLASH™ 3D NAND technology with PCIe Gen5 x4 interface, image credit SSSTC. TAIPEI, TAIWAN – Media OutReach Newswire – 19 August 2025 –Solid State Storage Technology Corporation (SSSTC) proudly announces the launch of the SSSTC CA8 Series, the world's first industrial M.2 SSD built with BiCS FLASH™ generation 8 3D flash memory technology and a PCIe® Gen5 x4 interface. Offered in the industry-standard M.2 2280 form factor, the SSSTC CA8 Series is available in 512 GB, 1 TB, 2 TB, and 4 TB capacities. The SSSTC CA8 Series complies with PCIe® 5.0 and NVMe™ 2.0, supporting ultra-high-speed data access with transfer rates up to 32 GT/s per-lane—doubling the throughput of PCIe® 4.0 interface. With SLC caching, the SSSTC CA8 Series achieves sequential read speeds of up to 14,000 MB/s and write speeds up to 12,000 MB/s, along with random performance of up to 2,000K IOPS (read) and 1,600K IOPS (write) — making it one of the fastest industrial SSDs on the market. At its core is KIOXIA's 8th generation BiCS FLASH™ 218-layer 3D TLC NAND Flash Memory with CMOS directly Bonded to Array (CBA) wafer bonding technology , enabling a 20% improvement in write performance, over 10% reduction in read latency, and up to 30% better power consumption. These enhancements make the SSSTC CA8 Series a critical enabler for data-intensive edge computing and high-performance industrial workloads. ADVERTISEMENT Designed for long-term reliability and broad system compatibility, the SSSTC CA8 Series features: • MTBF exceeding 3 million hours • Wide operating temperature range (0 °C to 85 °C) • Supply longevity of 5 to 10 years Its robust durability and consistent performance make it ideal for demanding applications such as AIoT, factory automation, networking, in-vehicle systems, and edge servers—especially those involving Edge AI workloads. To ensure data integrity and system resilience, the CA8 Series includes: • High-performance ECC • Power Loss Notification (PLN) to prevent data corruption during unexpected shutdowns • AES-256 encryption and TCG Opal support for advanced data protection As a subsidiary of KIOXIA, SSSTC is committed to delivering highly stable, high-performance SSD solutions to global enterprise and industrial customers. The SSSTC CA8 Series is scheduled to enter mass production in Q4 2025. For more information, please visit the SSSTC website Hashtag: #SSSTC The issuer is solely responsible for the content of this announcement. About Solid State Storage Technology Corporation (SSSTC) SSSTC, a subsidiary of KIOXIA—one of the world's leading memory solution providers—is dedicated to the development, manufacturing, and sales of solid-state drives (SSDs). By combining its in-house R&D capabilities with KIOXIA's cutting-edge NAND flash technology, SSSTC consistently delivers high-quality, high-reliability industrial-grade and enterprise-grade SSD solutions to customers worldwide. Trademark Notice: -NVMe is a registered or unregistered trademarks of NVM Express, Inc. in the United States and other countries. -PCIe is a registered trademark of PCI-SIG. -Other company names, product names and service names may be trademarks of third-party companies.
Web Release
6 hours ago
- Web Release
SANS report finds humans still the main attack vector as 80% of organizations flag social engineering as their number one risk
SANS report finds humans still the main attack vector as 80% of organizations flag social engineering as their number one risk The latest survey data from SANS Institute, the world's most trusted provider of cybersecurity training, reveals that 80% of organizations rank social engineering as the number one human-related risk—an already formidable threat now supercharged by AI. As attackers use artificial intelligence to craft more convincing and scalable deception tactics, the stakes for human error have never been higher. The data was a key insight from the 10th anniversary edition of SANS Institute's Security Awareness Report®: Embedding a Strong Security Culture. The report is based on SANS's largest survey ever, with feedback from over 2,700 security awareness practitioners from more than 70 countries who shared their unique perspectives to create the most comprehensive and revealing report yet. Lance Spitzner, Technical Director of SANS Workforce Security & Risk Training, highlights the report's significance on its 10th anniversary: 'The launch of the 10th edition of our Security Awareness Report is a major milestone for us and our most ambitious and far-reaching report to date. Designed as a dual-purpose playbook, it empowers security awareness professionals to not only drive organization-wide behavior and culture change but also advance their careers.' Key Findings and Insights · Top human risks: This year's data makes it clear: social engineering remains the top human risk by a wide margin (according to 80% of respondents), with phishing still leading, and smishing and vishing attacks growing in both frequency and sophistication. In a shift from last year's results, incorrect handling of sensitive data has now taken the second spot, followed by weak passwords and poor authentication. These changes reflect the evolving ways in which humans remain the primary attack vector, and why targeted, behavior-focused training continues to be essential. · Program challenges: Lack of time and staffing remain the two biggest challenges limiting industry professionals from building and managing an effective program. The report emphasises the use of tools like Generative AI to help security teams accelerate their impact at a global scale. · Benchmarking and maturity: For the sixth year in a row, the data confirms that larger security awareness teams drive more mature programs. On average, it takes at least 2.8 dedicated FTEs to meaningfully influence behavior—and four or more FTEs to begin shifting organizational culture. But staffing isn't everything. Sustained effort over time matters just as much. The longer your program has been in place, the more likely it is to be improving processes, strengthening partnerships and effectively engaging the workforce to reduce human risk. · Career development: In 2025, the average global annual salary for individuals working in security awareness is $116,091. In terms of geography, North America has the highest average annual salary at $129,961, almost identical to 2024's findings. In Europe, the average annual salary is $93,661. Spitzner concludes: 'This year's findings come against the backdrop of organisations facing rising threats like generative AI, deepfakes and other emerging threats. The report delivers timely, data-driven insights into how security teams are adapting, where gaps remain and which strategies are moving the needle. In a field where human risk is still under-reported, this report shines a spotlight on one of cybersecurity's most urgent challenges.' To read the full report and benchmark your program against industry standards, download the report here.
