Atempo is a France-based data protection company with data managing, moving and protection software products like Miria, Tina and Lina. Jeremy Smith tells us: “I’m the new President at Atempo and new primary shareholder.” Smith is the founder of HorizonH which buys “control positions in great European software companies and help them grow their revenue by helping them expand internationally.” Atempo CEO Luc D’Urso left in May, according to LinkedIn. There’s nothing about this in Atempo’s website.

…

Cloud storage provider **Backblaze **announced its B2 Cloud Storage platform’s strategic integration with Shareio, a new platform enabling creators to monetize and protect their digital content with robust security and clear economics. By combining Backblaze’s affordable, high-performance B2 Cloud Storage with Shareio’s decentralized monetization and digital rights management (DRM), creators can store their work cost-effectively while maintaining full control over distribution and monetization. Creators interested in Shareio’s monetization tools can sign up at www.shareio.com.

…

Broadcom announced an open, extensible ecosystem for VMware Cloud Foundation (VCF) enabling customers to build, connect, protect, and extend their modern private clouds. VCF 9.0 combines the agility and scalability of public cloud with the security, performance, architectural control, and total cost of ownership (TCO) benefits of an on-premises environment. VCF 9.0 delivers a unified platform for applications—traditional, modern, or AI—with consistent operations, governance, and controls. With VCF 9.0, customers gain a consistent operating model for the private cloud, spanning data centers, edge, and managed cloud infrastructure from service providers and hyperscalers. Get more information here.

…

Active data replicator Cirata announced a $6.7m, 3-year Data Integration software contract with IBM for a financial services customer, under IBM’s existing OEM sales agreement for the Big Replicate platform. This growth contract represents the largest contract with IBM in Cirata’s history.

…

Cloudera announced a platform update that integrates Trino, Cloudera Shared Data Experience (SDX), and Cloudera Octopai Data Lineage to deliver unified data access, control, governance, and lineage across the entire data estate. With this update, Cloudera customers can deploy Trino in data centers or public clouds and federate data across multiple systems using certified connectors. The integration with SDX unifies metadata and access controls, simplifying management and enabling secure, self-service data access. By unifying metadata, access controls, governance, and security across the entire enterprise, Cloudera delivers a single, secure endpoint to access all data—across multiple engines, clouds, and on-premises environments—without duplicating security or access policies. Additionally, Cloudera Octopai Data Lineage tracks the end-to-end journey of data, delivering transparency and trust for all data transformations, including data originating outside the Cloudera ecosystem.

…

Cohesity says in fiscal year 2025 (August 2024 – July 2025), its business with Microsoft expanded rapidly. It accelerated the adoption of Microsoft Teams, Azure, M365, GitHub, and the Microsoft Marketplace, while joint co-sell engagements increased more than 10 times year-over-year. There was global go-to-market efforts spanning North America, LATAM, Europe, Asia Pacific, and Japan, with a strategic partnership agreement and multi-year commitment to innovate, accelerate AI and data security solutions with joint customers. Cohesity reports nearly 200% growth in Microsoft Marketplace sales and over 10X year-over-year growth in co-sell engagements, jointly accelerating customer adoption of Microsoft Cloud and Cohesity

…

Cohesity research focused on looking at the material impact of cyber attacks on all aspects of business value, from stakeholder confidence, to ransomware payments impacts, as well as considerations around the impact of rapid AI adoption and compliance, found:

• Globally:

• 70% of publicly traded companies reported adjusting earnings or financial guidance after an attack.

• 68% said they observed an impact on their stock price.

• 73% of privately held firms redirected budgets from innovation and growth initiatives.

• 92% reported experiencing legal, regulatory, or compliance consequences, including fines, lawsuits, or other enforcement actions.

• Nearly half of UK organisations (43%) believe their cybersecurity strategy is watertight and requires little to no improvement. Complacency persists despite the spate of high-profile attacks on the UK’s retail sector this summer, which saw M&S’ profits more than halve

• Most UK businesses have had to rely on insurance to cover financial damage after an attack

• For a large majority, insurance did not adequately cover the cost of a cyberattack

• 71% of UK businesses have paid a ransom in the past 12 months

…

Azure-based backup, recovery and compliance supplier for Microsoft 365 CrashPlan has a deeper strategic collaboration with Microsoft. CrashPlan is introducing Microsoft 365 Archiving capabilities within its Backup and Recovery Platform. SharePoint Archiving capability automatically offloads inactive data to secondary storage while preserving user access through file stubs. By combining archiving, deduplication, compression, and incremental backup, organizations can significantly reduce storage expenses while maintaining full recoverability and compliance posture. CrashPlan and Microsoft are also exploring new ways to use AI-search to efficiently gain new insights from backup data. Learn more here.

…

Crusoe announced GA of Crusoe Managed Inference designed to run model inference, including large context and long-form text generation, on Crusoe Cloud with ultra-low latency, breakthrough time-to-first-token (TTFT) speed, and resilient scaling. AI developers can use Crusoe Managed Inference to rapidly deploy and automatically scale production-ready models, enabling new capabilities like AI agents and complex task automation. It’s powered by Crusoe’s proprietary inference engine, with MemoryAlloy technology, a cluster-wide KV cache that eliminates duplicate prefills by allowing GPUs to fetch prefix caches from local and remote nodes instantly. MemoryAlloy is a proprietary cluster-native memory fabric that enables persistent sessions, contextual continuity, and seamless scaling across an entire cluster.

…

Reuters says Databricks is in talks to raise capital at above $130 billion valuation, about 30% higher than its last financing round two months ago. It closed a $1 billion round in September, at a valuation of $100 billion.

…

**Datadobi **unveiled Advanced Storage Optimizer designed to take the guesswork out of managing enterprise storage. It delivers visibility into cost-reducing opportunities and is available in StorageMAP 7.4, the latest release of Datadobi’s platform. It identifies suboptimal use of storage tiers and enables users to model scenarios before migrations. It also identifies ageing and unused data, suggesting options to reduce costs, including what to archive, along with reports that can be shared with stakeholders. V7.4 also introduces enhanced support for Azure Blob archiving and new automated reporting features that simplify operations and integrate seamlessly with enterprise systems. These enhancements enable organizations to streamline data management and reinforce governance.

…

Andreas Dilger, the CTO of Whamcloud, **DDN’s **Lustre software unit, has left DDN and become a self-employed Principal Architect and Lustre Principal Architect for the The Lustre Collective (TLC), which is “dedicated to advancing the Lustre file system, the leading open-source, parallel distributed file system designed for large-scale cluster computing.” The funding basis is not known but may be from paid engagements with enterprises, research institutions, and cloud providers deploying Lustre-based systems. Focus areas include:

• Optimizing parallel I/O performance for HPC clusters.
• Enhancing Lustre for AI and machine learning workloads.
• Contributing to open-source Lustre projects and roadmaps.
• Providing expert consulting for Lustre deployments.
• Exploring integrations with emerging technologies.

…

DDN announced Sovereign AI Blueprints and NVIDIA-aligned Reference Designs, a portfolio of sovereign-by-design architectures that help governments, enterprises, and research institutions deploy AI at national and industrial scale. These architectures deliver >99% sustained GPU utilization, up to 40% lower energy use, and full data-control guarantees. AI at scale is as much a compliance and data-governance challenge as a performance one. As sovereign AI is expanding beyond governments to regulated industries, global, and cross-border organizations as the next biggest priority, these blueprints let organizations train sensitive data securely. This announcement is another step in DDN’s work to empower large-scale sovereign AI programs globally — from Yotta’s IndiaAI deployment to a partnership with Polarise to deliver next-generation sovereign AI clouds in Europe.

…

MeshDefend emerged from stealth with $2.3 million in pre-seed funding. It’s building next-generation intelligence for how enterprises run, protect, and scale their mission-critical data systems. The investment will support the development of MeshDefend’s AI-native Operating System for enterprise data infrastructure. Tejas Pandit, Co-founder & CEO, MeshDefend said, “We are on a mission to make every organization’s data infrastructure intelligent, autonomous, and secure. AI has advanced rapidly, yet data infrastructure operations still struggle for intelligence and context. MeshDefend aims to transform that reality by elevating the day-to-day of IT infrastructure teams, starting with backup and storage administrators. We are collaborating with infrastructure vendors, hyperscalers, MSPs, and GSIs to co-build impactful use cases on our Agent Mesh operating system.”

…

The Nutanix Cloud Platform will support Azure Virtual Desktop for hybrid environments. Benefits of this on Nutanix AHV will include:

• Hybrid Flexibility: Combine on-premises performance with Azure scalability for burst capacity and disaster recovery.
• Ability to run Microsoft applications: Native support for Microsoft 365, Teams optimisations, Entra, and security services.
• Optimised for Performance: Ideal for latency-sensitive and graphics-intensive workloads.
• Cost Efficiency: Leverage existing Microsoft licensing and Nutanix infrastructure for predictable economics.

…

M&E market focussed storage supplier OpenDrives is transitioning to a data services supplier with its Astraeus launch. Astraeus is OpenDrives’ cloud-native data services platform designed to unify, orchestrate, and manage data and applications across on-premises, cloud and hybrid environments. It delivers cloud agility and flexibility with on-premises performance and control, a unified view of all data regardless of location, reduced data sprawl, enhanced security and governance, and improved team productivity. As part of this rollout, OpenDrives is inviting organizations to join its Early Adopter Program (running until December 15, 2025) in its search for “lighthouse customers” to experience the platform firsthand and help shape its continued development. Get more info here.

Astraeus launch event.

…

Pliops announced a collaboration with Zilliz aimed at enabling affordable, large-scale Retrieval-Augmented Generation (RAG). This initiative targets multi-billion-scale vector databases at storage-level costs. Pliops and Zilliz addresses these challenges by combining the advanced vector database capabilities of Zilliz’s Milvus with the scale-out long-term memory of Pliops LightningAI. Together, these technologies enable efficient vector search, RAG, and large-scale context retrieval through hardware-accelerated KV Cache Offload – bringing scalable, cost-effective inference within reach for enterprise AI environments..

…

Pliops’ blogs address some of the biggest challenges in GenAI deployment and inference performance:

• Redefining KV-Cache Offload: How to enable cost-effective scaling for LLMs.
• Building Huge, Affordable Vector Databases: A look at using memory-efficient ANN algorithms, GPUs, and shared KV storage for massive vector DB performance.
• Beyond Traditional Caching: Introducing persistent memory tiers that unlock long-term memory for next-gen inference.
• Setting the Standard for Multi-Turn Benchmarking: Best practices for evaluating real-world LLM workloads with vLLM.

Pliops announced a Pliops x DapuStor Partnership delivering optimized KV-cache acceleration for GenAI workloads, and a New Collaboration with J-Tech expanding high-performance LLM acceleration solutions into the Korean market.

…

Vector search supplier **Qdrant **launched Tiered Multitenancy, tackling one of the biggest operational challenges in vector search: managing thousands of small tenants alongside just a few extremely large ones in a single collection—eliminating latency issues, index sprawl, and routing complexity. All tenants start in a shared collection, and high-traffic tenants can be promoted to dedicated shards with a single API call—no reindexing, downtime, or client-side routing changes. It’s the first model that combines tenant isolation and global cross-tenant search inside one vector database, eliminating noisy-neighbor effects while keeping architectures simple. The result: consistent, high performance, lower infrastructure costs, improved security and a scalable model that fits real-world production RAG and Agentic AI environments.

…

Mark Cree is the new President for Acumera and Scale Computing, which was bought by Acumera in July. Scale Computing co-founder and CEO Jeff Ready has resigned from the company.

…

Seagate launched new Exos JBODs, the 4U100 and 4U74 systems with SAS-4 connected SAS or SATA HAMR disk drives. the 4U100 has100 x 32 TB drives in a 4 RU chassis storing up to 3.2 PB plus 70% more efficient cooling and 30% lower power use than previous SAS-3 generations. The 4U74 has, of course, 74 drives. The 4U100 is longer, at 11460.0mm (45.12 in) vs the 4U74’s 932.0mm (36.69 in) standard rack depth, hence the larger number of drives. The Exos 4U100 and 4U74 JBODs will be available next quarter through Seagate’s authorized channel partners worldwide. Download the datasheet here.

SEagate Exos JBODSs

…

VDURA went well over the marketing top again. It says the world just witnessed data history in the making. At Supercomputing 2025 (SC25), VDURA and Phison Electronics teamed with Icelandic strongman Hafþór “The Mountain” Björnsson to achieve the fastest data transfer ever recorded.

Before a packed audience at Booth 2033, Björnsson transferred 136.5 petabytes (PB) of data in just 27 seconds, hitting a sustained rate of 5.1 PB per second (≈ 303 PB per minute). The demonstration showcased the combined power of Phison Pascari enterprise SSDs and VDURA’s Flash-optimized parallel data storage infrastructure, unleashing bandwidth once thought impossible in AI and HPC environments. “VDURA is built for moments like this,” said Ken Claffey, CEO of VDURA. “What Thor achieved on the show floor symbolizes what our customers achieve every day. Extreme performance, scalability, and durability, all working together at record-breaking speed.” Who said storage can’t be fun?

VDURA and Phison strong man.