Intelligent ETL supplier Adeptia has hired three execs: Tim Bond, Chief Product Officer; Jill Ransome, Vice President of Marketing; and John Moore, Vice President of Revenue Operations. Charles Nardi, CEO of Adeptia, said: "Our leadership will be instrumental as we continue helping organizations transform fragmented data into trusted, AI-ready information that powers automation, operational intelligence, and business growth.”
Ransome will lead global marketing strategy and execution, expanding Adeptia's market presence and driving awareness of the company's Intelligent ETL approach to enterprise data automation; Bond will oversee product strategy and innovation, advancing the company's AI-native platform that the foundation organizations rely on to build, orchestrate and scale their data workflows; and Moore will lead sales focused on scaling go-to-market performance, improving operational efficiency, and supporting customer growth.
…
Source connector Airbyte says its v2.1 SW improves scalability, deployment flexibility, governance, observability, and operational reliability, which helps organizations manage increasingly complex data pipelines that can be used for AI applications. We're told it delivers:
Improved performance and scalability for large enterprise deployments;
Expanded enterprise governance and operational controls;
Enhanced self-managed deployment capabilities;
Additional reliability and observability improvements for production pipelines;
Expanded support for modern AI and analytics architectures;
Connector and synchronization improvements across structured and unstructured data sources.
Airbyte 2.1 builds on the company’s strategy of becoming the operational data movement layer for modern AI infrastructure, enabling enterprises to continuously connect trusted operational data to cloud data platforms, vector databases, large language models (LLMs), and AI agents. For more information about Airbyte 2.1, visit: Airbyte 2.1 Release Notes.
…
**AMD **is acquiring MEXT, a maker of AI-driven memory optimization technology. MEXT Predictive Memory is a software that makes flash appear as DRAM to the operating system. The result is a new price-performance tier that delivers DRAM-class performance at flash-level costs and capacity. MEXT begins by off cold memory pages from DRAM to a 50X lower-cost tier, flash. The MEXT Predictive Memory Engine uses AI to continually predict which memory pages in flash are likely to be requested by the application, and moves them back to DRAM— before the application even requests them. From the application's perspective, all relevant memory pages are always findable in DRAM, maintaining performance.
Through this AI-powered prediction, MEXT says it can keep application performance intact within a much smaller DRAM footprint - yielding substantially lower costs. Alternatively, MEXT can instantly enable 2-4X memory capacity on existing systems (through flash-as-memory), yielding superior performance for memory-bound workloads.
…
Data orchestrator and manager Arcitecta has promoted SW engineer Ewen Smith to managing director of Arcitecta Europe. It’s hired Nina Mangold from GRAU DATA to be its European channel sales manager. Smith has experience spanning engineering, enterprise technology and complex data environments. Smith will lead the company’s European growth strategy and customer engagement across the region.
Mangold brings more than 18 years of experience in data management and archiving, with a background serving the science, research and big data communities. Her industry network is expected to support Arcitecta’s continued momentum with scientific research institutions and GLAM organizations (galleries, libraries, archives, and museums), including customers such as the Imperial War Museums.
…
Arcitecta, GRAU DATA, and COMBACK are showcasing a combined system for end-to-end data intelligence and archives at ISC 2026. By integrating Arcitecta’s Mediaflux research data management platform with GRAU DATA’s Metadata-Hub/XtreemStore and COMBACK’s BDT ORION enterprise tape infrastructure, they’ve bridged the gap between active data management and metadata-driven archiving into a single, scalable architecture. For more information, visit: https://www.arcitecta.com/events/2026/isc/.
…
Data trust business **Ataccama **announced a native integration with Databricks to bring data quality, lineage, and governance signals from enterprise source systems, including SAP, Oracle, and mainframe environments, into the Databricks platform, helping customers build AI applications and agents on trusted data. Available now in the Databricks Marketplace, the integration connects Ataccama’s governed quality signals and metadata context directly to Databricks pipelines, workflows, and AI tools through the Ataccama MCP Trust Layer.
…
Cirata announced it’s entered into a revised Original Equipment Manufacturer agreement with IBM. Under the terms of the Revised OEM Agreement, Cirata Symphony - the Company's Data Orchestration product - will be offered by IBM to its customer base and branded as Cirata Symphony for IBM Big Replicate. It will provide a unified control plane for petabyte-scale data movement, operation, and management.
…
Startup Clario, developing SW to eliminate enterprise data ROT (Redundant, Obsolete and Trivial), launched from stealth with $6 million in seed funding. The round was led by Preface Ventures, with participation from Foster Ventures, Golden Sparrow, High Sage Ventures, Moment Ventures, Mentors Fund, Page One Ventures, Rain Capital, Ridge Ventures, and Transform VC, as well as angel investors Michael Callahan and Baris Aksoy. 78 percent of all stored enterprise data is unstructured, and conservative industry estimates suggest more than a third is effectively garbage.
Clario connects directly into existing file and content systems or unstructured data environments, including Google Drive, SharePoint, OneDrive, Box, and Confluence, to scan metadata and surface all garbage. When a file or article is flagged, Clario activates intelligent workflows via Slack and Teams, prompting customers to keep, archive, or delete the data – and only gets paid when customers make one of these decisions. Over time, Clario learns from these decisions to build an autonomous system that routinely cleans, organizes, and maintains unstructured data.
…
Data protector CloudCasa by Catalogic announced CloudCasa Disaster Recovery (CloudCasa DR) for Kubernetes powered by HPE Alletra Storage MP B10000. It brings enterprise class disaster recovery to Kubernetes applications and KubeVirt virtual machines across HPE and hybrid cloud environments. Building on CloudCasa's existing support for backup and recovery of HPE Kubernetes Service (HKS) clusters, the new integration adds orchestrated disaster recovery with near-zero data loss and fast recovery times. Rather than restoring from backups, CloudCasa DR uses Alletra Storage MP B10000 native replication.
Protected workloads include both Kubernetes applications and KubeVirt virtual machines such as those running on Red Hat OpenShift Virtualization and SUSE Virtualization. Supported platforms include HPE Kubernetes Service, Red Hat OpenShift, SUSE Rancher, and other Kubernetes environments. Customers can also use HPE StoreOnce and HPE X10000 Object Storage as backup storage targets for retention. …
Nearly three-quarters (72%) of global IT leaders say a lack of real-time data infrastructure is stalling their efforts to scale AI, according to a new 2026 Data Streaming Report from Confluent.
…
Genie9 has made BigMind DR Local free — block-level whole-machine imaging, cross-hardware bare-metal restore (offline driver injection), and ransomware-canary detection, with backups on the user's own disk or NAS, free for up to 5 users. Their cloud platform, BigMind Resilience, offers per-seat DR plus ransomware detection. Read more here.
…
HPE has updated its HPE AI Factory with Nvidia, using Vera CPUs. Nvidia Agent Toolkit software, including Nemotron open models, NemoClaw, and the OpenShell secure runtime, provide an agent operating system that lets customers monitor agent behavior, enforce policies, and reduce deployment risk. New HPE Zerto Software capabilities help customers identify when rogue agent actions take place and use continuous data protection to rewind to a clean slate. HPE Private Cloud AI also supports secure local agent registration, providing customers with the ability to approve AI models, skills, and tools while adhering to centralized governance and security policies.
With HPE Alletra Storage MP X10000, customers can automatically apply metadata and governance policies to prepare data for AI applications and cut token response times by up to 20X. HPE Data Fabric Software broadens data availability for agentic workflows by extending support of model context protocol (MCP) to Apache Airflow and introduces an enterprise AI inventory that adds metadata to distributed data. A standalone HPE Data Fabric appliance, available on HPE ProLiant Compute servers, is intended to simplify and accelerate deployment.
…
**Lexar **is now looking to break the M.2 2280 form factor mold, aiming to usher customers into a new era of compact storage with its mSSD (micro Solid State Drive) standard. Its first mSSD product, the Play X SSD, delivers read speeds up to 7,400 MB/s and write speeds up to 6,500 MB/s by operating as a PCIe Gen 4 drive, in about half the size of a regular SSD. I2 has an M.2 2280 slot adapter. According to the company, the mSSD unifies the controller, NAND, and PMIC in a single module, bringing more stability, system integration, and the ability to scale this product for enterprise use, as it could be deployed in businesses to help build servers, AI platforms, and other complex scenarios, thanks to its custom controller and firmware design.
The mSSD is a lot more durable and has built-in data protection compared to other major SSDs. Its unified architecture and integrated design bring better dust, moisture, and physical shock resistance. Read more here.
…
Memory-focussed semiconductor supplier Netlist has initiated new legal proceedings against Samsung in the U.S. International Trade Commission (ITC) and the U.S. District Court for the Eastern District of Texas. These actions are based on the infringement of Netlist’s U.S. Patent Nos. 12,646,537 and 12,650,937, which respectively read on Samsung’s high-bandwidth memory (HBM) products and Samsung’s DDR5 RDIMMs and MRDIMMs. The new ITC complaint names Google, Supermicro, Nvidia and Broadcom as additional respondents. Netlist is seeking exclusion and cease and desist orders against Samsung and other respondents.
…
Open-source enterprise Postgres company ** pgEdge** announced pgEdge ColdFront, a transparent data tiering setup for PostgreSQL. Unlike other alternatives, ColdFront’s cold tier is fully writable: UPDATE and DELETE work on archived rows through the same SQL the application already uses, with no code changes and no rehydration required. We're told older data moves automatically to Apache Iceberg in Parquet format on any S3-compatible object store at up to 90 percent lower storage cost. Meanwhile, the complete dataset stays readable and writable through a single Postgres table name, and cold-tier scans operate at analytical speed thanks to the DuckDB vectorized columnar engine.
…
Quantum has filed a Form 12b-25, Notification of Late Filing, with the Securities and Exchange Commission with respect to its Annual Report on Form 10-K for the fiscal year ended March 31, 2026 ("Fiscal 2026”). The Company intends to file its Annual Report on Form 10-K as soon as practicable, and in any event on or before June 29, 2026, the fifteenth calendar day following the original due date, as provided under Rule 12b-25. Fiscal 2026 is also the first year in which the Company's financial statements have been audited by its current independent registered public accounting firm, which has involved additional first-year audit procedures. The Company does not currently expect the delay to affect its previously announced preliminary unaudited financial results for the fourth quarter of Fiscal 2026 or its financial statements for any prior fiscal period.
…
SK Hynix has shipped samples of 12-stack HBM4E, next-generation DRAM for AI, to major customers.
…
Snowflake announced Unlimitail - the omnichannel retail media company backed by Carrefour and Publicis - has chosen Snowflake to power its retail media network. This collaboration will reach 250 million shoppers across Europe and Latin America, enabling privacy-preserving, data driven advertising across Unlimitail's network of 120 retail sites.
…
BeeGFS supplier ThinkParQ launched BeeOND Enterprise (BeeGFS On Demand), a standalone subscription-based edition that gives AI and HPC teams the complete on-demand storage pipeline: local NVMe across compute nodes dynamically pooled into a high-performance scratch file system, spun up in seconds at job start and torn down at job end, plus BeeGFS Copy, the parallel data staging engine that stages input data into the scratch tier before a job begins and writes results back to global storage on completion, ensuring BeeOND’s NVMe speed benefits the full job pipeline, not just the computational phase. It says this addresses one of the most overlooked inefficiencies in high-performance computing: the terabytes of fast local NVMe storage sitting idle in every compute node while organizations continue to purchase additional dedicated burst-buffer appliances to address storage bottlenecks.
BeeGFS Copy is a parallel data staging engine that moves data at scale between multiple persistent data lakes and the BeeOND scratch tier. It stages input data into the scratch tier before a job begins and writes results back to global storage on completion, ensuring BeeOND’s NVMe speed benefits the full job pipeline.
…
The Swiss National Supercomputing Centre (CSCS), part of ETH Zurich, presented two technical papers at the 2026 Cray User Group (CUG) conference in Nice, France, highlighting VAST Data’s role in advancing high-performance infrastructure for modern AI and HPC workloads. CSCS presented research on * “Architectural Isolation for Sensitive Workloads: Enabling Trusted Research Environments on HPE Cray EX Systems,*” which received the CUG 2026 Best Paper Award. CSCS also presented
“Evaluating the Operational Viability of VAST Data as a Scratch File System for Alps,” which examined the use of VAST infrastructure for HPC and AI workloads on Alps, CSCS’s HPE Cray EX supercomputer.
Both papers presented by CSCS at the Cray User Group will be available in the next months from the official website.
…
Australian Neocloud SharonAI and **VAST Data **announced an expanded strategic partnership under which Sharon AI will deploy 600PB of the VAST AI Operating System across its AI cloud infrastructure. The commitment makes the VAST AI Operating System the foundational data layer for Sharon AI's sovereign AI infrastructure, serving government, enterprise, research, and AI-native customers across Australia and Asia-Pacific. This deployment represents one of the largest and most advanced sovereign AI data foundations in the Asia-Pacific region.
…
**Veeam Software **announced a major expansion of its global strategic alliance with Everpure, marking a series of joint innovations and unveiling the next generation of integrated cyber resilience and DataAI Resilience. Veeam and Everpure are turning alliance momentum into an integrated resilience roadmap across Everpure’s Enterprise Data Cloud (EDC) and Kubernetes, combining anomaly-driven workflows available today, fleet-scale EDC integration planned in Veeam Data Platform v13.1, and Cyber Resilience Delivered as-a-Service through trusted partners to reduce operational complexity and improve recovery confidence at enterprise scale.
…
**Veeam Software **research shows a growing data disconnect across EMEA organizations. While 99 percent of enterprise decision-makers agree that data sovereignty is critical, at the organizational level, the majority are actively deprioritizing it in favor of accelerating AI. The result: AI workflows have become the region’s biggest data visibility gap, with 40 percent of leaders calling “data used for AI or analytics” their top operational blind spot.
…
Morgan Stanley analysts met with WD CEO Irving Tan and CFO Kris Sennesael. The analysts noted that management hinted at the potential to use [HAMR] IP for applications beyond Hard Disk Drives.
…
Wiwynn demo’d an Nvidia SCADA (SCaled Accelerated Data Access) 6 RU server at Computex 2026. The server has Nvidia's Vera CPU, 4 x RTX Pro 6000 Blackwell graphics cards, 4 x PCIe 6.x switches, and 4 x ConnectX-9 SuperNIC cards, along with 96 liquid-cooled E3.S format SSDs, cooled by six separate cold plate modules. Using Micron 9650 Pro drives, with 30.72 TB capacity, the server has 2.949 PB of capacity. With SCADA, the GPUs initiate and control storage I/O operations and the data path. It has an aggregated 4K random read speed of 528 million IOPS
…
AT TOKYO Corporation and Zadara announced a strategic partnership to launch the AT TOKYO region of Zadara Cloud at AT TOKYO data centers. The service, branded as ATBeX Edge Compute powered by Zadara, is scheduled to be available in August 2026. It’s designed to support growing demand for AI infrastructure and sovereign cloud capabilities in Japan. The service combines Zadara’s Federated Edge Cloud platform with AT TOKYO’s ATBeX network fabric to deliver integrated compute, storage, and networking while enabling domestic data residency and high connectivity.