We’re Tracking the Future of Security

Category: Data Centers

  • Meta Tests Robots From Watney, Kinova and ABB to Automate Data Center Maintenance

    Meta Tests Robots From Watney, Kinova and ABB to Automate Data Center Maintenance

    Meta is piloting robots from three vendors, Watney Robotics, Kinova and ABB, to handle physical maintenance tasks inside its data centers, including swapping network cables, power-cycling servers, reseating components and inspecting equipment, according to an August 31, 2026 report from WIRED based on current and former employees familiar with the trials.

    What the Robots Are Doing

    At Meta’s Altoona, Iowa campus, a pair of dual-armed Watney robots has been tested on cabling work since June 2025, supervised by human operators and still slower than a person, per the report. At the Prometheus campus in New Albany, Ohio, four-wheel ABB robots equipped with a scissor-lift riser and a six-axis arm are being used to reseat hardware components, and Meta is separately evaluating a Kinova Gen3 robotic arm for power-cycling servers. Meta has also deployed simpler robots that remotely restart devices by physically pressing power buttons. Kinova and ABB declined to comment to WIRED, and Watney did not respond to the outlet’s requests for comment.

    Why It Matters

    One Meta data-center worker told WIRED that a working cable-swapping system could eventually take on as much as 80% of some technicians’ current workload, though that figure is an employee estimate rather than a company-published target, and the robots still struggle with dense cabling, tight corners and tasks that require sustained autonomy. As AI-driven data center buildouts accelerate, the trials point to facility operations and physical security converging with the same automation trends reshaping server hardware itself.

  • Texas Governor Orders Pause on New Data Center Approvals Amid Grid Interconnection Strain

    Texas Governor Orders Pause on New Data Center Approvals Amid Grid Interconnection Strain

    Texas Governor Greg Abbott has ordered a pause on approvals for new large data center projects seeking to connect to the state’s power grid, a move that Houston Public Media reported on August 27, 2026 could delay roughly 300 large data center projects, though not every planned facility in the state is covered by the pause.

    The order responds to mounting pressure on the Electric Reliability Council of Texas (ERCOT) interconnection queue, as AI-driven data center demand has produced a wave of large-load requests competing for grid capacity and connection timelines. According to the report, Beth Garza, who previously served as ERCOT’s independent market monitor, said the interconnection process could take significantly longer than originally projected, whether because of the governor’s pause or because the initial timeline itself was unrealistic; Garza said she “will be pleasantly surprised” if by April 2027 the industry has a clear picture of which data center loads will ultimately be able to move forward.

    Texas has emerged as one of the largest hubs for hyperscale and AI-focused data center construction in the country, drawing investment from major cloud and AI infrastructure providers seeking access to relatively fast permitting and available land, but also straining grid planning as operators request power commitments that can rival the demand of entire cities. The pause reflects a broader tension states are navigating between courting large-scale data center investment and protecting grid reliability and consumer electricity costs for existing residential and industrial customers.

    For the physical security and critical infrastructure sector, large-scale data center buildouts have significant downstream implications beyond power supply: campus perimeter security, access control, and fire and life-safety systems are typically scoped and budgeted alongside the underlying facility and power infrastructure, meaning delays or restructuring of a project’s grid interconnection timeline can directly affect the pace of associated physical security procurement and installation work tied to new builds.

  • Nvidia Makes Minority Investment in Data-Center Power Developer Cloverleaf

    Nvidia Makes Minority Investment in Data-Center Power Developer Cloverleaf

    Nvidia has made a minority investment in privately held Cloverleaf Infrastructure, a company that arranges power and site infrastructure for AI data-center projects across the United States, the companies said on Friday, August 21, 2026, according to Reuters.

    Chipmaker Moves Further Upstream Into Power

    Financial terms of the investment were not disclosed, but the Wall Street Journal reported the same day, citing people familiar with the deal, that Nvidia was expected to invest up to several hundred million dollars. Cloverleaf works with utilities, energy providers and investors to secure power and other infrastructure for data-center sites, and the company says it has delivered multiple gigawatt-scale projects across North America since its founding in 2024. As part of the arrangement, Cloverleaf will deploy Nvidia’s DSX platform to help optimize decisions on site selection, power, cooling and computing infrastructure for the data centers it develops.

    J.P. Morgan Securities served as exclusive financial advisor and Kirkland & Ellis as legal counsel to Cloverleaf in structuring the deal, according to trade publication POWER.

    Part of a Broader Pattern of Financing Data-Center Buildout

    The Cloverleaf investment came just days after Nvidia announced a separate $1.5 billion investment in SoftBank-owned SB Energy to support the PORTS-Pike technology campus, a 10-gigawatt, OpenAI-linked data-center project in Pike County, Ohio. Nvidia CEO Jensen Huang said of that earlier deal that “AI is becoming infrastructure — the foundation for intelligence in every industry — and land, power and shell have become vital.” Taken together, the two deals illustrate how Nvidia has begun taking a more direct role in financing and developing the data centers that ultimately buy its AI computing systems, rather than simply supplying chips to third-party developers. Power availability, rather than chip supply, has increasingly become the binding constraint on how quickly new AI data-center capacity can come online.

    Sources

  • Nvidia Customers Reportedly Warned of AI Server Price Hikes as Groq Deal Moves Forward

    Nvidia Customers Reportedly Warned of AI Server Price Hikes as Groq Deal Moves Forward

    Nvidia’s largest server customers have been told to expect price increases of more than 15% in many cases for AI servers, driven by rising memory chip costs, Bloomberg reported over the weekend of August 22–23, 2026, with the story continuing to be discussed heavily in tech and financial media on August 24. Separately, CNBC reported that Nvidia said racks built with hardware from Groq — in which Nvidia recently made a roughly $20 billion related investment — will come online later this year.

    What is driving the increases

    According to Bloomberg’s reporting, the price increases are being driven primarily by soaring memory chip costs rather than by Nvidia’s own component pricing, as server makers pass through higher costs for the memory needed to build AI-optimized systems. The report landed the week Nvidia is scheduled to report quarterly earnings, adding to investor focus on demand signals for the company’s AI hardware.

    Why it matters

    Rising AI server costs affect every organization planning large-scale data center buildouts, including security, video analytics and AI-driven monitoring platforms that increasingly depend on GPU-accelerated infrastructure. Higher per-server costs can slow the pace at which cloud providers and enterprises expand AI compute capacity, even as demand signals — including Nvidia’s push to bring Groq-linked infrastructure online this year — suggest that AI infrastructure investment is continuing at a rapid pace despite the added cost pressure.

    Sources

    More coverage like this is available on Technology News.

  • Data Center Physical Security: A Layered Design Guide

    Data Center Physical Security: A Layered Design Guide

    Data centers are among the most security-sensitive facilities in modern infrastructure. They contain high-value equipment, critical data services and dependencies that support banking, telecom, cloud platforms, government systems and enterprise operations. Physical security must therefore be designed as a layered system.

    Layer 1: Site Boundary

    The outer boundary should discourage casual access and provide early detection. Depending on the site, this may include fencing, vehicle barriers, perimeter cameras, thermal imaging, radar or fiber-optic intrusion detection. The goal is to create enough distance and warning time before a person reaches the building.

    Layer 2: Vehicle and Visitor Control

    Vehicle gates, intercoms, license-plate recognition and visitor-management systems establish accountability before entry. Delivery vehicles and contractors should follow workflows different from permanent staff.

    Layer 3: Building Access

    Access control should use strong credentials, anti-passback logic and role-based permissions. High-security sites may add biometrics, mantraps or multi-factor physical authentication. Credentials should be linked to HR and identity-management processes so access changes when employment status changes.

    Layer 4: White Space and Critical Rooms

    Server halls, network rooms, power systems and storage areas require additional zoning. Not every employee who can enter the building should be able to enter every technical space. Door events should be correlated with video so investigations can reconstruct who entered, when and under which authorization.

    Video and Analytics

    Cameras support verification, investigation and compliance. Coverage should focus on entrances, corridors, cages, loading areas and critical equipment zones. Analytics can help identify tailgating, unusual movement or occupancy patterns, but should supplement rather than replace access-control logic.

    Environmental and Fire Protection

    Physical security also includes resilience. Aspirating smoke detection, thermal monitoring, leak detection, clean-agent suppression and power-system monitoring protect availability from non-criminal threats.

    Cyber-Physical Security

    Security devices themselves are networked computers. Cameras and controllers need firmware management, segmentation, strong credentials and logging. Compromised physical-security devices can create both cyber and physical risk.

    Conclusion

    The strongest data-center design uses multiple independent layers so failure of one control does not expose the asset. Perimeter security, identity, video, environmental monitoring and cybersecurity should all contribute to a single risk-based architecture.

  • Clean-Agent Fire Suppression for Data Centers: Design Guide

    Clean-Agent Fire Suppression for Data Centers: Design Guide

    Data centers concentrate electrical equipment, energy, cooling infrastructure and business-critical services into spaces where even a small fire can create disproportionate operational loss. Clean-agent suppression is designed for environments where rapid extinguishment and minimal residue are priorities.

    What is a clean agent? Clean agents are gaseous fire-suppression media that leave little or no residue after discharge. Depending on the technology, suppression may be achieved through heat absorption, chemical interaction with the flame process, or reduction of oxygen concentration within safe design limits.

    Why data centers use them Water remains an essential fire-protection tool, but uncontrolled water exposure can damage servers, storage and electrical distribution. Clean-agent systems can suppress a developing fire without coating equipment in powder or liquid residue. They are therefore commonly considered for server rooms, network rooms, control rooms and other high-value electronic spaces.

    Detection matters as much as suppression The most effective design starts with early detection. Aspirating smoke detection can identify incipient smoke before conditions become severe. A staged alarm sequence can verify the event, alert operators, stop selected ventilation systems and initiate the discharge logic.

    Room integrity and pressure relief A gaseous system only performs as intended if the protected enclosure can retain the required concentration for the specified period. Door gaps, cable penetrations and ventilation openings can reduce performance. Enclosure integrity testing and pressure-relief design are therefore critical parts of commissioning.

    Not a substitute for an overall fire strategy Clean-agent systems should sit inside a broader architecture that includes detection, compartmentation, emergency power procedures, portable extinguishers, possible sprinkler protection and documented recovery plans.

    For data-center owners, the engineering objective is not simply to extinguish fire. It is to limit downtime, protect people, preserve critical infrastructure and make recovery predictable. The best clean-agent design is therefore one that integrates suppression with detection, HVAC control, electrical isolation and business-continuity planning.

  • NVIDIA and Financial Partners Target $500 Billion for AI Infrastructure

    NVIDIA and Financial Partners Target $500 Billion for AI Infrastructure

    August 10, 2026 — NVIDIA announced partnerships with six financial firms intended to mobilize more than $500 billion of third-party capital for AI infrastructure over time.

    What happened

    NVIDIA said independent financing platforms with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR were designed to mobilize more than $500 billion of third-party capital. The announcement describes a target over time; it is not the same as NVIDIA placing $500 billion into a completed fund or guaranteeing immediate chip purchases.

    Why it matters

    The structure is significant because it treats AI compute and supporting infrastructure as a financeable asset class. That could broaden access to capital while adding questions about utilization, collateral value, operating risk and the durability of demand.

    Security and infrastructure impact

    Security and resilience affect the value of financed infrastructure. Lenders and operators will need credible controls for asset tracking, facility access, cyber risk, downtime, insurance and recovery across long equipment and financing lifecycles.

    Sources

    ← Back to Technology News

  • Alphabet Raises 2026 AI Capital Spending Guidance to $205 Billion

    Alphabet Raises 2026 AI Capital Spending Guidance to $205 Billion

    July 22, 2026 — Alphabet raised its full-year 2026 capital expenditure guidance to a range of $195 billion to $205 billion, up from a prior $180 billion to $190 billion range, citing accelerating AI infrastructure demand.

    What happened

    Alphabet’s chief financial officer announced the increase on the company’s second-quarter 2026 earnings call. Roughly half of the spending is allocated to servers, with about 40 percent going to data centers and networking equipment.

    Why it matters

    Combined 2026 AI-related capital spending across Alphabet, Microsoft, Amazon and Meta is now projected in the $600 billion to $725 billion range, a scale of infrastructure investment with direct implications for power grids, construction supply chains and the physical and cyber security of the resulting facilities.

    Security and infrastructure impact

    SectechMedia continues to track how this data-center boom translates into demand for perimeter security, fire and life-safety systems and OT/IT convergence expertise at newly built AI campuses.

    Sources

    ← Back to Technology News

  • AI Data Center Growth Expands Power, Safety and Security Risk Planning

    AI Data Center Growth Expands Power, Safety and Security Risk Planning

    2025–2026 — Rapid AI data-center construction is forcing operators to treat power availability, fire safety, physical protection and cyber resilience as one connected risk program.

    What happened

    AI workloads are accelerating demand for high-density computing facilities and the energy systems that support them. The International Energy Agency’s Energy and AI work describes the growing relationship between data centers and electricity systems, while large infrastructure announcements show that the buildout is already moving into physical delivery.

    Why it matters

    The risk is not simply higher electricity consumption. Dense compute changes cooling, backup power, battery, fire-detection and business-continuity requirements. New sites also add construction-stage access, supplier and commissioning risks before normal operations begin.

    Security and infrastructure impact

    A defensible design connects cyber and physical controls: segmented building and OT networks, managed contractor identities, layered perimeter detection, validated fire scenarios and a command process that can correlate facility, security and IT alarms.

    Sources

    ← Back to Technology News

  • AMD and Anthropic Plan Up to 2 Gigawatts of MI450 GPU Deployment

    AMD and Anthropic Plan Up to 2 Gigawatts of MI450 GPU Deployment

    July 22, 2026 — AMD and Anthropic announced a strategic partnership for up to two gigawatts of MI450-series GPU capacity, with the first gigawatt planned for 2027.

    What happened

    AMD and Anthropic said they plan to deploy up to two gigawatts of AMD Instinct MI450 Series GPUs in Helios rack-scale systems. Their official statement places the beginning of the first gigawatt in the first half of 2027, so the announcement is a forward deployment commitment rather than completed capacity.

    Why it matters

    The agreement signals that frontier AI developers are seeking large-scale alternatives and complementary supply to established accelerator platforms. It also shows that future model capacity is being negotiated in power-scale terms, not only in numbers of chips.

    Security and infrastructure impact

    Multi-gigawatt plans have consequences for grid interconnection, cooling, construction, fire safety and site protection. Procurement teams should separate announced maximum scope from contracted phases and verify delivery milestones before treating capacity as operational.

    Sources

    ← Back to Technology News