Containers

Storage news ticker – April 7

Chris Mellor Blocks & Files editor
Published

Backblaze announced that one of its original Storage Pod 1.0 chassis will be preserved in the Computer History Museum's permanent collection. The donated unit, internally known as "ul010," was the 11th storage pod in production and stored customer data for years. Read more here.

…

Backblaze's Q1 2026 Performance Stats report reveals significant performance variation across providers and regions, with different architectures excelling under different conditions. The full report, including detailed charts, throughput benchmarks, regional comparisons, and complete testing methodology, is available here.

…

Catalogic's CloudCasa has joined the Nutanix Kubernetes Platform (NKP) Partner Catalog, bringing Kubernetes-native backup, recovery, disaster recovery, and migration capabilities to NKP users. CloudCasa in the NKP Partner Catalog supports the broader NKP Anywhere vision by helping customers not only run Kubernetes wherever they need it, but also protect, recover, and migrate workloads across those environments with greater confidence.

…

Object storage supplier Cloudian says it has just completed a survey of 212 senior IT decision-makers – all at organizations where 76 percent already have more than half their workloads in public cloud. These aren't cloud skeptics. And yet: 

• 89 percent plan to expand their on-premises footprint in the next two years 

• 75 percent have already moved some workloads back from public cloud in the past 24 months 

• 84 percent are over their cloud storage budgets – nearly 1 in 5 by more than 30 percent

• Just 0.5 percent report being under budget 

• 99 percent say data sovereignty is at least a moderate factor in infrastructure decisions 

Cloudian asserts that the story isn't "enterprises souring on cloud." Experienced cloud operators, running workloads at scale, are being more precise about where the model works and where it breaks down. Egress fees, compliance complexity, and AI latency requirements are forcing a reckoning that the original cloud economics never priced in. Get the report here.

...

Databricks is partnering with STATION F, described as the world’s largest startup campus, located in Paris, France, to help European startups better use data and AI. The campus occupies a historic 1920s railway freight depot known as Halle Freyssinet. It houses up to 1,000 startups and thousands of entrepreneurs under one roof. Databricks will open a dedicated office at STATION F offering training and coaching, ecosystem programs, technical sessions, and customer and partner events.

…

Egnyte AI Assistant screenshot.
Egnyte AI Assistant screenshot

Cloud-based collaborator Egnyte announced two major additions to the Egnyte Content Cloud: AI Safeguards, giving organizations granular control over how AI interacts with sensitive content, and a deeply integrated AI Assistant that acts as a built-in collaborator across Egnyte workspaces. AI Safeguards enables IT and compliance teams to precisely define which users, groups, file locations, and file properties can be processed by AI. AI interactions are visible and auditable through detailed reports, helping reduce the risk that AI becomes a black box inside the organization.

Egnyte says its AI Assistant evolves its copilot into a true built-in collaborator. It isn't a chatbot layered on top of your content; it understands context, including each user's content, permissions, and business context. Users can ask questions, generate content, and automate tasks directly within their work in Egnyte. Egnyte also announced:

  • AI in the Desktop App allows users to ask questions, access agents, and trigger workflows directly from the desktop, where they spend most of their time. According to Egnyte's internal data, more than 70 percent of content across all Egnyte domains comes from the desktop app, which is the preferred experience for users working on complex, mission-critical content such as CAD designs and Adobe projects.
  • AI Actions surfaced throughout Egnyte experiences enable users to leverage intelligent insights without switching between documents or applications or leaving the product.
  • Agentic Search that goes beyond keyword matching to understand natural language questions to analyze intent, content, and metadata, and return direct, cited answers, helping users find information.
  • AI Audit Reporting provides full visibility into AI use across the organization, tracking who analyzed which content and when, so policies can be enforced with confidence.

…

Everpure stated: "Due to sustained, industry-wide demand and elevated component costs, Everpure is adjusting prices effective March 30, 2026. Most of our portfolio will see an Average Selling Price (ASP) increase of approximately 40 percent, while Evergreen//One will be limited to an increase of approximately 10 percent on longer-term contracts. Our priority remains protecting customer economics while ensuring consistent supply and support." 

…

AI-focused data orchestrator Hammerspace announced support for FIPS 140-3 validated cryptography, enabling its data platform to meet US government standards for cryptographic security and support deployments across federal, defense, healthcare, and financial services environments. Integration into the Hammerspace Data Platform is planned for the end of 2026.

…

The Keepit Annual Data Report 2026 highlights:

  • 9 in 10 enterprises have validated bulk recovery, demonstrating maturity in their disaster recovery preparedness 
  • Identity systems are tested four times less often than productivity systems, even though losing identity access can prevent access to all other SaaS applications. 
  • 90 percent of restores are single file downloads, indicating that simple data loss incidents are most common, and that IT administrators appreciate the ability to do granular, immediate recovery. 
  • High-profile global outages – including major cloud and security incidents – did not lead to increased recovery testing, suggesting that even visible disruptions rarely prompt validation of recovery readiness. 

 Get a copy here.

…

Nvidia has invested $2 billion in custom semiconductor designer Marvell and announced an agreement to connect Marvell to the Nvidia AI factory and AI-RAN ecosystem through NVLink Fusion. Marvell will provide custom XPUs and NVLink Fusion-compatible scale-up networking, while Nvidia will provide the supporting technologies, including Vera CPU, ConnectX NICs, Bluefield DPUs, NVLink interconnect and Spectrum-X switches, and the rack-scale AI compute.

The companies say they'll attempt to transform the world's telecommunications networks into AI infrastructure with Nvidia Aerial AI-RAN for 5G/6G, and advance world-class networking for AI, including advanced optical interconnect solutions and silicon photonics technology.

…

SAP has agreed to acquire Reltio and its master data management (MDM) software to help customers make their SAP and non-SAP enterprise data AI-ready. Terms of the deal were not disclosed. Once closed, we're told the acquisition will strengthen SAP Business Data Cloud (SAP BDC) – integral to SAP's AI-First and Suite-First strategy – and accelerate the evolution of SAP BDC to a fully interoperable enterprise data platform for enterprise-wide agentic AI.

…

The SNIA has confirmed that SDC: StorageAI in Denver, Colorado, will take place on April 29. This one-day (SNIA Developer Conference) event will explore how AI is exposing challenges in data services, particularly in how storage, networks, and compute must work together to meet new performance demands. The agenda features execs on the front lines of the biggest trends shaping the AI landscape. Registration is $149.

Breakout presentations include:

…

The UK IBM Storage Scale User Group has confirmed dates for the 2026 meeting. It will run a dedicated New User Day ahead of the main conference, aimed at those newer to Storage Scale or looking to strengthen their foundational knowledge. Both events are now open for registration via Eventbrite: New User Day here and main conference here.

...

Veeam was named Customers' Choice in Gartner's 2026 Peer Insights Voice of the Customer: Backup and Data Protection Platforms. Veeam received an overall customer rating of 4.8 out of 5 based on 124 reviews as of December 31, 2025, with 98 percent of customers willing to recommend Veeam's solutions, 73 percent awarding Veeam five stars, and 27 percent four stars.

…

VergeIO, a VMware migration target, announced support for NVIDIA RTX vWS and vPC on the latest vGPU 20 release. This enables customers to deploy graphics-intensive professional applications from cloud-based virtual desktops. VergeOS abstracts GPU management tasks, including driver provisioning, resource allocation, active monitoring, and Nvidia Multi-Instance GPU (MIG) configuration. VergeIO says its initiative brings automated GPU management, intelligent provisioning of right-sized virtual GPUs, and near-bare-metal performance to private cloud environments while giving customers confidence in a validated, NVIDIA-backed deployment path. 

 …

Nutanix and VMware alternative Virtuozzo reckons its Virtuozzo Infrastructure v7.3 software has new features that help service providers and IT teams build more efficient and scalable cloud services. Some of the highlights include:

  • GPU metering – track usage in GPU-hours and enable monetization of GPU workloads 
  • VM migration visibility & control – full tracking, history, and easier troubleshooting 
  • S3 object tagging – improve data organization and enable automation 
  • S3 multi-tenancy – project-based isolation aligned with compute environments 
  • Storage & monitoring improvements – better performance and new Grafana dashboard for replication visibility 

…

Walrus has introduced a MemWal plugin for NemoClaw/OpenClaw – agent orchestration frameworks for building autonomous AI workflows – that allows agents to store durable memory on Walrus and retrieve it across runs, environments, and teams without building custom storage infrastructure. The MemWal plugin connects NemoClaw and OpenClaw agents to MemWal's memory model so developers can add persistent memory with minimal integration work., the company says. 

… 

RERAM developer Weebit Nano completed an A$87 million ($56 million) capital raise, which was oversubscribed, and launched a Share Purchase Plan expected to raise an additional A$15 million ($6.5 million). CEO Coby Hannoch said: "We completed this raise not because we had an immediate need for cash, but because we are at a unique moment where multiple vectors are aligning. We have an opportunity to significantly improve our ReRAM to the point where it is the clear leader in the market, enabling us to take significant market share, as well as establish a leading position in the In-Memory Compute (IMC) domain. The funds raised will enable us to capitalize on these opportunities." 

He believes AI is becoming foundational in embedded systems, with designers focused on delivering AI at the edge within strict power and area constraints. This is driving interest in architectures such as in-memory computing (IMC), where ReRAM can play an important role over time.

egnyte weebit nano containers veeam