Vertiv

Recent news

• Nov 4, 2026

  Modular AI Data Centers Gain Ground as Crusoe Debuts Unit

  Crusoe has announced its sixth-generation modular data center platform, Spark Liquid, for AI compute workloads.

  • The new Spark Liquid unit is designed to support up to 3.5 MW of compute capacity and is built for Nvidia GB300 NVL72 and Vera Rubin NVL72 racks; Crusoe says it delivers five times the compute density of earlier Spark generations.
  • Production will join Crusoe’s existing Spark line in Brighton, Colorado, in 2027; the platform uses closed-loop direct liquid cooling and is aligned with Nvidia’s DSX POD modular blueprint.
  • Crusoe says Spark units are built in its 352,000-square-foot factory in Brighton and can be brought online in as little as a few months from a ready site; the company acquired Easter-Owens Electric in 2022 to bring manufacturing in-house.
• Oct 22, 2026

  Nvidia Unveils DSX POD Amid Greenpeace Protest

  Nvidia has announced DSX POD, a prefabricated modular reference design for AI data centers, at its GTC event in Berlin.

  • Nvidia said the factory-built, pre-integrated modules could cut deployment time from about six months to as little as 14 days and are intended for partners including Dell Technologies, HPE, Schneider Electric, and Vertiv.
  • Jensen Huang said a gigawatt-scale data center costs $50 billion and that a three-month delay could mean $10 billion in lost revenue; the article also describes Greenpeace protesters interrupting the keynote and raising concerns about water use, electricity bills, and environmental impacts.
• Oct 21, 2026

  Nvidia Unveils DSX Pod for Faster AI Data Center Builds

  Nvidia has announced DSX Pod, a containerized AI data center module, at GTC Berlin.

  • The module includes seven Vera Rubin NVL72 racks plus networking, cabling, and cooling already installed, and Nvidia says it can reduce deployment time at a prepared site from about six months to as little as 14 days.
  • Nvidia said it is working with 17 partners, including Bull, Dell Technologies, HPE, Lenovo, Eaton, Schneider Electric, and Vertiv, with availability expected in early 2027; the company also outlined manufacturing expansion plans and introduced the OpenShell Reference Architecture for AI governance and security.
• Sep 24, 2026

  Nvidia selects three BESS solutions for new AI facilities qualification programme

  Nvidia has announced Nvidia DSX Ready, a new qualification programme for products and solutions that meet its reference design requirements for AI factory infrastructure.

  • The programme launched earlier this week on 21 September and initially includes battery energy storage systems (BESS) and cooling distribution units (CDUs) as qualified categories.
  • First listed BESS providers are Hitachi Energy, LG Energy Solution, and Tesla; first CDUs are from LG Electronics, LiquidStack, and Vertiv. Nvidia said qualification does not replace site-level engineering or imply site-level stability.
• Aug 28, 2026

  AI Rack Density’s Real Limits: Power, Cooling, Failure Risk

  The article analyzes how AI is pushing data center rack density higher and argues that cooling, power delivery, and grid access now set the main limits on AI rack design.

  • It cites Uptime Institute data showing modal rack density rose to 11 kW in 2026 from 9 kW in 2025, while Nvidia GB300 NVL72 reaches up to 142 kW per rack and Rubin Ultra NVL576 “Kyber” is specified at roughly 600 kW for H2 2027.
  • It says direct-to-chip liquid cooling now handles 100-150 kW per rack and is the dominant liquid method; 800 VDC is shipping with Vera Rubin NVL72, and vendors including Vertiv, Schneider Electric, Eaton, and Delta have commercial 800 VDC offerings slated for H2 2026. It also references Foxconn’s 40 MW Kaohsiung-1 facility in Taiwan, the Open Compute Project Mount Diablo specification finalized in March 2026, and Microsoft/Meta hardware demonstrations in July 2026.
• Jul 7, 2026

  Generate Capital's Sven Semmelmann joins OpenAI as head of compute capital markets

  OpenAI has announced that Sven Semmelmann will join as head of compute capital markets.

  • Semmelmann is joining OpenAI from Generate Capital, where he spent six years as head of structured finance; he will lead capital markets strategy for AI infrastructure and support OpenAI’s expanding global compute platform.
  • OpenAI CFO Sarah Friar said Semmelmann will develop the financing solutions and partnerships needed for OpenAI’s infrastructure plans; the company also noted it has about $1.4 trillion in compute-related spending commitments over the next eight years and has filed a confidential draft registration statement for a potential IPO.
  • The article also references Generate Capital’s investments in Soluna, a partnership with Vertiv, support for esVolta, and a bad investment in Pine Gate that led to layoffs, but these are background details rather than the main announcement.
• Jul 2, 2026

  AWS in-row heat exchanger to reduce water use by 9% over evaporative air-cooled data centers

  Amazon has detailed new sustainability measures for AWS and its global data centers in its annual sustainability report.

  • The company said its In-Row Heat Exchanger (IRHX) captures heat directly from high-density AI hardware at the rack, and that adding more AI racks with IRHX should deliver a 9% reduction in water use versus evaporative air-cooled data centers once fully operational.
  • Amazon said it is expanding liquid-to-chip cooling in both new and existing data centers, scaling smart-metering to detect leaks, and reported that its data centers used 2.5 billion gallons of water in 2025 while reducing cooling water use by 938 million liters.
  • The report also said 31 data centers had on-site water treatment, 26 used reclaimed water for cooling, 14 had rainwater collection systems, and Amazon has contracted 13 utilities to supply reclaimed water for 130 data centers, saving 6 billion liters of fresh water annually.
• Jun 12, 2026

  From Components to AI Factories: Peter Panfil Says the Future of Data Centers Is All About Integration at Scale

  Vertiv’s Peter Panfil presented a keynote at the 2026 7x24 Exchange Spring Conference outlining a vision to treat data centers as integrated “AI factories” that prioritize execution velocity, factory-assembled high-density modules, and outcomes-oriented metrics.

  • Main announcement: Vertiv (Peter Panfil) advocated for integrated, factory-produced HAC “hacks” as repeatable building blocks to accelerate deployment and reduce on-site integration. He noted a rapid module evolution from ~1.5 MW (a year ago) to ~6 MW current modules, with discussions already underway around 12 MW configurations; modules are now being assembled and fully tested in factories (including fluid charging and capacity validation) to enable plug-and-play site installation and faster commissioning.
  • Supporting details: Panfil proposed replacing traditional efficiency metrics with tokens-per-dollar-per-watt (debated/refined to tokens per watt per dollar), emphasized behavioral modeling/digital twins (example: coolant excursion reduced from ~9°C to ~3°C with modest buffering), and highlighted energy strategies including BYOP/on-site generation, UPS smoothing for grid stability, and community acceptance measures (waste heat reuse, grid support).
• Jun 9, 2026

  Vulnerabilities discovered in Trane, Vertiv data center products

  Claroty’s Team82 has identified vulnerabilities in Trane Tracer SC+ HVAC controllers and Vertiv Liebert IS-UNITY-DP network cards.

  • Main announcement: Claroty (Team82) released research reporting that unauthenticated attackers could disrupt HVAC and UPS/power management operations, potentially causing service degradation, outages, or hardware damage; Vertiv’s Liebert IS-UNITY-DP vulnerabilities were scored 9.8 CVSSv3 and Vertiv recommends Liebert RDU101 and IS-UNITY firmware updates, while Trane recommends updating Tracer SC+ to v6.3 or later.
  • Background / details: The Tracer SC+ issues involve multiple unauthenticated API routes exposing device and downstream device information (including BACnet and LonTalk-connected devices) enabling reconnaissance, lateral movement, or direct manipulation; Vertiv network card flaws could allow attackers to execute arbitrary code or issue an ‘output OFF’ command to shut down UPS-powered devices; both vendors have been informed and published mitigations in Team82 reports.
• May 26, 2026

  How Power Electronics Cut Generator Run Hours in AI-Scale Data Centers

  DatacenterKnowledge reports that advanced power electronics and battery systems can substantially reduce diesel-generator run hours at AI-scale data centers while preserving uptime.

  • Main announcement/action: DataCenterKnowledge outlines a shift from a “diesel-first” resilience model to a diesel-last approach using grid-forming inverters, AI-capable UPS, high-voltage DC (800 VDC) architectures, and Universal Damping STATCOMs; specific reported capabilities include ON.energy’s 3.5 MW AI UPS units (reported ~3 GW in operation or construction) offering 1–8 hours of backup, and Dimaag.ai’s 800 VDC Zenius + Zifer system presented to ERCOT (April 24, 2026) for evaluation.
  • Background and details: The article references a 2025 Virginia grid fault with an aggregate loss of about 1.5 GW that triggered generator activations; Ramboll’s Universal Damping (UD) STATCOM is proposed as a software-upgrade approach to damp oscillations; Eaton’s Power Xpert 9395P with EnergyAware was tested at Microsoft’s Innovation Center in Boydton, Virginia and evaluated by PJM; Vertiv’s EnergyCore Grid BESS is UL9540A-tested for mission-critical use.
• May 12, 2026

  Cowtown Angels Portfolio Co. Strategic Thermal Labs Acquired by Ohio-Based Vertiv

  Vertiv has acquired Georgetown-based Strategic Thermal Labs (STL), integrating STL’s chip-level liquid-cooling expertise into Vertiv’s thermal-chain strategy.

  • Acquisition announced: Vertiv (NYSE: VRT) acquired Strategic Thermal Labs (STL); STL contributes cold-plate design, server-side liquid cooling, and high-density thermal validation capabilities to strengthen Vertiv’s thermal-chain interface for AI and high-performance computing environments. Cowtown Angels (TechFW) participated in STL’s April 2024 seed round (led by Carrier Ventures; included Cowtown Sidecar Fund-I). Article date: May 12, 2026.
  • Background & details: STL was founded in 2014. Vertiv said the deal improves its ability to simulate and emulate high-density compute conditions, optimize thermal and power-train interaction, and support customers across design, integration, commissioning, and lifecycle operations. Quotes included from Caitlin McMinn (TechFW) and Scott Armul (Vertiv).
• May 12, 2026 · Outsourcing Today

  Romania can become the “new Norway” in the data center industry. The strategic window for attracting major AI investments is 12 months

  Tema Energy and DataCenter Forum Romania participants announced that Romania has a ~12-month window to attract AI and hyperscale data center investments and can provide up to 500 MW almost immediately.

  • Main announcement/action:Tema Energy and forum speakers declared a strategic opportunity: Romania can leverage uncongested grid zones to offer up to 500 MW almost immediately and has roughly 12 months to attract major AI/hyperscale investments; an initial anchor project of >100 MW (example: ClusterPower in Craiova) was cited as a potential catalyst, with realistic project implementation timelines of 18–20 months.
  • Background and details: The forum (DataCenter Forum Romania 2026) featured speakers from NVIDIA and Accelerated Infrastructure Capital (AIC); data centers currently represent 0.2% of Romania’s electricity use (vs 6% Netherlands, 24% Ireland). The article referenced Norway’s Stargate Narvik (~230 MW renewable target, ~100,000 NVIDIA GPUs) as a comparative model and emphasized needs such as liquid cooling, renewables + storage, and rapid grid connections.
• May 5, 2026

  New Data Center Developments: May 2026

  Data Center Knowledge published a monthly roundup highlighting global data center project announcements, regulatory moves, and investment commitments driven by hyperscale AI demand.

  • Main announcement: The roundup catalogs multiple concrete project actions including Aligned Data Centers’ Project Caprock (540 MW, 313-acre campus in Hale County, Texas; initial delivery Q1 2027), EdgeCore’s completion of $1.5 billion in financing for two Northern Virginia hyperscale centers, and Yondr Group energizing a 27 MW Toronto facility expected in mid-2026. It also notes major investment commitments such as Digital Realty’s near S$7 billion Singapore plan (S$4.3 billion for new data centers) and AWS increasing planned investment in Mississippi to $25 billion.
  • Context and details: The piece outlines parallel regulatory updates in U.S. states (Maine vetoed a moratorium; Wisconsin revised We Energies tariff rules; North Carolina advanced legislation to require hyperscalers to cover infrastructure costs), workforce and partnership initiatives (Equinix Foundation with ODATA, Cisco, Vertiv launching training in Brazil, cohorts mid-2026), and other regional projects and financings (TikTok €1 billion Finland site; Ark Data Centres >€600 million Barcelona project; Equinix land purchases in South Africa totaling ZAR 890 million).
• Apr 28, 2026 · IT Brief Asia

  Vertiv acquires Strategic Thermal Labs in cooling push

  Vertiv has announced it has acquired Strategic Thermal Labs.

  • Main announcement: Vertiv has acquired Strategic Thermal Labs to expand its engineering capabilities in liquid cooling for dense computing systems, adding expertise in cold-plate design, server-side liquid cooling, and thermal validation; the business will support customers through design, integration, commissioning and ongoing operations.
  • Background and details: The acquisition strengthens Vertiv’s thermal-chain strategy to manage heat from chips through supporting infrastructure; Vertiv is based in Westerville, Ohio and operates in more than 130 countries. The company said the transaction will not change its support for open, interoperable infrastructure and aims to improve simulation/emulation of dense compute conditions.