Director Legal - Energy & Power Procurement

Nvidia builds European 'sovereign' AI clouds with partners. In a period in which demand for local computing power and artificial intelligence is soaring, Nvidia is rapidly expanding its network of European cloud partners. Cloud infrastructure providers in Europe, Scandinavia, Africa and Central Asia are currently building so-called AI factories on a large scale based on the latest Nvidia DSX and Vera Rubin architectures. The initiative is intended to provide AI developers and industry with 'sovereign' computing capacity for training, fine-tuning and deploying models, close to the location where the data and expertise are located. The network of regional cloud partners gives companies and researchers more control over where data is stored and how their AI applications are managed. A major challenge in scaling these AI factories is energy consumption. Cloud providers such as Nscale use the Nvidia DSX platform to optimize power consumption. In tests with Nvidia GB300 NVL72 systems, the so-called MaxLPS technology enabled Nscale to deploy 37 percent more GPUs within the same energy budget and achieve almost 50 percent more inference throughput. Nscale also plans to roll out the new Vera Rubin infrastructure in five European markets. Modular data center builders such as Crusoe are also joining with liquid-cooled AI factories ('Spark') based on the Vera Rubin platform. This will make it possible to offer low-latency AI services close to European end users. Other regional providers expanding their Nvidia infrastructure include Aleria, Cassava Technologies, Firebird, G42, Radiant, SWI Compute and Vultr. Open models and industrial applications. In addition to hardware capacity, the European cloud partners enable developers to work with open AI models and libraries. For example, cloud provider Nebius integrates technologies such as Nvidia Groq 3 LPX and Vera Rubin NVL72 on its Nebius Token Factory platform to generate AI tokens faster. Nebius also offers a 'Physical AI Workbench' on which robotics companies such as Rhoda AI, RoboForce and Sereact build their applications. French AI developer Mistral, in turn, is expanding its own accelerated computing infrastructure in European data centers with 3,800 Nvidia Grace Blackwell GPUs to support its latest multimodal model, Mistral Large 4. Delivery of the successor Vera Rubin infrastructure at Mistral is scheduled for 2027. In the United Kingdom, Fireworks and Era4 are collaborating on the rollout of liquid-cooled data centers based on Blackwell Ultra GPUs for applications in the financial sector, insurance and healthcare. Built-in security for autonomous AI agents. With the rise of autonomous AI agents that can independently perform actions within business systems, the need for strict security frameworks is growing. At GTC Berlin, Nvidia presented the OpenShell Reference Architecture, an integrated security model supported by numerous European cloud partners. The architecture combines hardware and software to set clear boundaries on what an AI agent can and may access: * Nvidia Vera Rubin NVL72 processes the AI models and responses. * Nvidia Vera CPUs with OpenShell provide the agent's isolated working environment (harness) and tool use. * Nvidia BlueField-4 DPUs with 'Sentry' software enforce the configured security rules directly at the network hardware level. Partners such as CoreWeave, DigitalOcean, Lambda, Lightning AI, Nebius, Nscale and Vultr support this OpenShell architecture. In this way, the parties aim to ensure that companies can safely put autonomous AI agents into production without the risk of unauthorized access to sensitive business data. About Witold Kepinski. Witold Kepinski (1969) is Director, Editor-in-Chief and Director Content of Dutch IT Channel and Dutch IT Leaders. Witold Kepinski has been active in the IT media and tech business sector for 25 years. With a specialized team of editors, bloggers and video makers, Witold Kepinski provides insight into tech business trends and applications with which IT decision-makers and channel partners make an impact.

Anthropic in $45B Nscale deal for Vera Rubin AI cloud. Anthropic nears $45B Nscale deal for AI computing powered by Nvidia Vera Rubin chips. October 19, 2026 Here is the HTML code for the cryptocurrency news article, structured as requested. - Advertisement - * Anthropic is negotiating a $45 billion computing power agreement with Nscale for a new West Virginia facility. * The facility will run on NVIDIA's next-generation Vera Rubin chips and generate 460 megawatts of capacity. * Nscale is targeting a $30 billion valuation ahead of its IPO after a $3.36 billion pre-IPO funding round. Anthropic is in advanced discussions to acquire roughly $45 billion in computing power from Nvidia-backed cloud infrastructure firm Nscale, according to reports. The prospective agreement centers on a second facility at Nscale's Monarch data center campus in West Virginia. The new building is slated to generate 460 megawatts of computing capacity powered by Nvidia's next-generation Vera Rubin chips. An agreement could materialize in the coming days. The West Virginia campus was initially slated to house infrastructure for Microsoft, which signed a preliminary letter of intent in March. However, Microsoft withdrew from the site, clearing the way for Anthropic. Anthropic already has existing infrastructure commitments with Nscale valued at approximately $44.6 billion, according to regulatory filings. The creator of the Claude AI model has rapidly emerged as a dominant consumer of global data center capacity. - Advertisement - Nscale recently closed a $3.36 billion pre-IPO financing round led by Dan Loeb's Third Point with participation from Nvidia, Apollo Global Management, Citadel, Hudson Bay Capital, and the Abu Dhabi Investment Council. The company is targeting a $30 billion valuation for its public debut. Anthropic is also weighing its own public offering as it scales its computational infrastructure to meet surging enterprise demand. Retail sentiment on Stocktwits was 'bullish' with 'high' message volumes for Anthropic. Stay in the loop. Join 1.2K Smart Traders News October 20, 2026 Threat actors are exploiting a chain of four vulnerabilities in Plex Media Server, dubbed... News October 20, 2026 Anthropic will commit 5% of its shares to a new Public Benefit Fund ahead... News October 19, 2026 China's CSI 300 index recorded its 10th consecutive weekly decline, losing 3.4% last week... News October 19, 2026 Shareholders of Nasdaq-listed Super League approved a deal giving Japanese Bitcoin treasury firm Metaplanet... News October 19, 2026 Bitcoin trades at $82,164, holding above a key support line after touching $83,040, as... If you're looking to dive into the world of blockchain technology, you're in for a treat. The field is rapidly evolving and the potential...

Microsoft pulls out of UK Nscale data-center, shaking SaaS cloud backbone. SaasRise - Oct 19, 2026 Microsoft has withdrawn from a joint data-center project with UK neocloud firm Nscale, citing power-supply constraints. The move stalls a high-profile AI-focused facility and forces SaaS companies that rely on Azure UK capacity to rethink their infrastructure roadmap. Why it matters. The withdrawal highlights a critical, often overlooked, dependency of cloud-based SaaS on reliable power infrastructure. As AI workloads become more compute-intensive, any shortfall in electricity supply can directly curtail the ability of SaaS firms to deliver real-time, AI-enhanced services. This creates a strategic imperative for operators to diversify cloud footprints and embed energy-risk assessments into product roadmaps. Furthermore, the episode may accelerate policy discussions around grid modernization in the UK. If major cloud providers encounter repeated power-supply roadblocks, regulators could be pressured to prioritize renewable and high-capacity energy projects, which would ultimately benefit the broader SaaS ecosystem by stabilizing the underlying infrastructure needed for next-gen applications. Key points. * Microsoft ends UK data-center partnership with Nscale over unresolved power-supply constraints. * The stalled project was slated to host a large AI-focused facility, part of a global $7 trillion data-centre investment wave. * SaaS companies relying on Azure UK must consider multi-cloud redundancy or alternative capacity sources. * Four major cloud providers plan $725 billion in AI-related capex, a 77% YoY rise, intensifying demand on local grids. * The incident may prompt UK policymakers to accelerate grid upgrades to support AI-driven cloud services. Analysis. Microsoft's retreat from the Nscale venture is less a corporate whim than a symptom of a structural bottleneck: energy supply. The AI arms race has turned data-centres into power-hungry behemoths, and the UK's grid, still transitioning to renewable sources, is struggling to keep pace. For SaaS operators, this translates into a new risk vector that sits alongside the usual concerns of churn, net-retention, and competitive pricing. Companies that have built their go-to-market on a single-cloud, AI-native proposition now face a potential choke point that could erode service reliability and inflate unit economics. Historically, cloud providers have mitigated such risks through geographic diversification - spreading workloads across regions and even continents. The current scenario forces a re-evaluation of that playbook: European SaaS firms may need to adopt a true multi-cloud stance, not just for vendor lock-in avoidance but for energy resilience. This could accelerate the adoption of hybrid-cloud architectures, where on-premise or edge compute supplements public-cloud capacity during peak demand. In the longer term, the episode could reshape the competitive landscape. If Microsoft's power-supply challenges persist, rivals like AWS and Google Cloud may seize market share by securing more robust energy contracts or by investing directly in renewable generation. SaaS founders should watch for emerging pricing differentials and potential incentives tied to green-energy usage, as providers look to differentiate on sustainability as well as performance. The strategic takeaway is clear: infrastructure risk, once a background concern, is now a front-line factor in SaaS growth planning.