Getting your news
Attempting to reconnect
Finding the latest in Climate
Hang in there while we load your news feed
IBM
Data center news, project activity, and monthly briefings for IBM.
Editor's picks
-
AI factories explained: Building, deploying and improving AI at scale
Telehouse Canada announced a major infrastructure upgrade including first-of-its-kind direct liquid-to-chip cooling in collaboration with Enwave to scale AI workloads and bring compute closer to end users.
- Main announcement: Telehouse Canada has introduced a major infrastructure upgrade that includes direct liquid-to-chip technology (delivered in collaboration with Enwave) to enable sustainable cooling, low-latency performance, and support for high-density AI deployments in a metro environment; the upgrade is described in Telehouse’s recent press release linked in the article.
- Background and details: The article cites IBM data that 42% of enterprise-scale companies (>1,000 employees) report active AI deployment; Telehouse’s upgrade reportedly removes up to 80 per cent of heat from high-power server components, reducing reliance on power-intensive CRACs and server fans, and is positioned to support private/hybrid deployments where sensitive data cannot be hosted in public clouds.
-
PFAS phase-out and liquid cooling: What US data center operators must do now
The Environmental Protection Agency (EPA) has designated PFAS as “forever chemicals” and under TSCA has delayed but maintained reporting requirements, with the reporting deadline pushed to January 31, 2027.
- Main announcement: The EPA/TSCA reporting requirement for PFAS use affects data center operators who must report PFAS use once requirements take effect; the EPA has extended the reporting deadline to January 31, 2027 after three prior extensions. The article frames this as a compliance shift that is already halting growth in two-phase immersion cooling and encouraging migration to PFAS-free options.
- Background and details: The piece notes 3M has phased out Novec, several U.S. states (New Jersey, Maine, Massachusetts, Minnesota, Washington) have reporting or ban plans, and market data shows single-phase DTC holds a 55% market share in 2026, using a 75% water / 25% glycol coolant; Schneider Electric (with Motivair) is promoting PFAS-free single-phase DTC solutions.
-
A faster way to estimate AI power consumption
MIT researchers and the MIT-IBM Watson AI Lab have developed EnergAIzer, a fast estimator that predicts power consumption of AI workloads on GPUs.
- Main announcement: The team (led by Kyungmi Lee with senior author Anantha P. Chandrakasan) unveiled EnergAIzer, a lightweight estimation model that produces reliable GPU power estimates in a few seconds (vs. hours or days for traditional emulation) and achieved about 8 percent error when tested on real GPU workloads; the research is being presented at the IEEE International Symposium on Performance Analysis of Systems and Software.
- Background and details: The method leverages repeatable software optimization patterns to generate fast estimates, includes correction terms derived from real GPU measurements for accuracy, can be applied to future/emerging GPU configurations, and the work was funded in part by the MIT-IBM Watson AI Lab.
-
Q.ANT Brings Commercial Photonic Computing to the United States, Appoints Bruno Spruth as CTO
Q.ANT has announced the opening of its U.S. headquarters in Austin, Texas and the appointment of Bruno Spruth as Chief Technology Officer.
- Main announcement: Q.ANT has opened its U.S. headquarters in Austin, Texas, appointed Bruno Spruth as CTO, and plans to increase U.S. headcount to 20 employees over the next six months; the company also plans to localize chip manufacturing as it brings commercial photonic processors to North America and will operate a Native Processing Server as a PCIe co-processor in existing data centers.
- Background and details: Q.ANT closed a USD80 million Series A from investors including Duquesne Family Office, Hermann Hauser, Cherry Ventures, UVC Ventures, L-Bank, imec.xpand, Verve Ventures, Grazia Equity, EXF Alpha, LEA Partners, Onsight Ventures, and TRUMPF; it deployed the first commercial photonic processor in production at LRZ in 2025 and claims NPUs deliver up to 30x energy efficiency and 50x performance versus conventional processors. The article references U.S. companies committing over $690 billion to AI infrastructure in 2026 as market context.
-
IBM Announces Strategic Collaboration with Arm to Shape the Future of Enterprise Computing
IBM has announced a strategic collaboration with Arm to develop dual‑architecture hardware that helps enterprises run AI and data‑intensive workloads with greater flexibility, reliability, and security.
- Collaboration scope: IBM and Arm will explore virtualization technologies to allow Arm-based software environments to operate within IBM enterprise platforms (including IBM Z and LinuxONE), enable enterprise systems to recognize and execute Arm applications, and address reliability, security, and data sovereignty requirements. The announcement references IBM hardware platforms such as the Telum II processor and Spyre Accelerator as part of IBM’s ongoing hardware investment.
- Ecosystem and implementation: The partnership is focused on long-term ecosystem growth by creating shared technology layers to broaden software ecosystems and increase workload portability; the work targets high-availability operations, security, and performance/efficiency for AI and data-intensive workloads. No specific timelines or financial commitments were disclosed in the release.
-
DataBank Adds Two New Board Directors with Deep Expertise in Data Center Infrastructure and Technology
DataBank announced the appointment of two new members to its Board of Directors: Rob Johnson (former Vertiv CEO) and Daniel Sturman (former Roblox CTO).
- Main announcement: DataBank appointed Rob Johnson and Daniel Sturman to its Board to support expansion of its high-density data centers and low-latency interconnection hubs; the company serves 2,500+ enterprise, hyperscale cloud, and AI customers, operates 65+ data centers in 25+ markets and 20 interconnection hubs, and guarantees 100% uptime.
- Background and details: Johnson served as Vertiv CEO (2016–2022) and holds roles at G2 Ventures, Neos Partners, and Rehlko; Sturman served as Roblox CTO (2020–2024) and oversaw growth from ~100M to nearly 400M monthly users. Tom Yanagi of DigitalBridge commented on his 10-year support of DataBank’s strategy. Press assets and contact links (LinkedIn, YouTube, corporate site, phone 1(800) 840-7533) are provided in the release.
Recent news
-
7x24 Spring Conference: Future-Proofing the AI Data Center Amid New Bottlenecks, New Risks, New Rules of Execution
The 7x24 Exchange Spring Conference highlighted that the data center industry must “future-proof” people, processes, and governance while introducing the development of a new quality framework, DCE 9000, through the Telecommunications Industry Association (TIA).
- Main announcement/action: The conference sessions, led by speakers including Sol Rashidi, Google’s Govind Ramu and Gino Tozzi, and panelists from Victaulic, T5 Data Centers, DLB Associates, and Oracle, emphasized future-proofing operational and human systems for AI-scale data centers and noted that DCE 9000 is being developed through TIA as a common quality management framework for data center infrastructure equipment, suppliers, contractors, and operators.
- Background and details: Presentations flagged concrete operational risks: roughly 70% of organizational change initiatives fail (research cited by Google), urgent needs around liquid cooling commissioning, contamination control, supplier governance, documentation discipline, and workforce development, and recommended earlier involvement of chemical treatment/water-quality specialists and tighter integration across engineering, contractors, commissioning providers, equipment manufacturers, and IT teams.
-
IBM Targets Enterprise AI Control Layer with ‘Operating Model’ Push
IBM announced an ‘AI operating model’ built around data, agents, automation, and hybrid infrastructure to move enterprises from isolated AI projects to coordinated systems embedded in core business processes.
- Main announcement: IBM introduced an AI operating model and positioned Watsonx Orchestrate as a control plane and Concert as the operational remediation layer; the company announced an integration of Watsonx.data with Confluent’s streaming technologies to enable real-time data pipelines and live context for agent-driven systems.
- Background and details:Arvind Krishna framed the shift citing over 70% of enterprise data remaining on-premises; analysts (Tekonyx, HyperFrame Research, Moor Insights & Strategy) warned enterprises are fragmented on data and lacking production governance, and IBM emphasized hybrid/on-premises deployment rather than building foundation models or competing on hyperscale infrastructure.
-
How do you feel about AI’s environmental impact?
Gannon Knight’s Roundtable asked Gannon students and staff their thoughts on AI’s environmental impact.
Main announcement/action: Gannon Knight surveyed Gannon University participants — including Sam Mason, Ph.D. (Director of Project NePTWNE; oversees the Center for Lake Erie Education and Research, CLEER), Zara Toomaney (senior, Environmental Engineering), and Sophia Cuzzola (junior; IBM SkillsBuild student ambassador) — on April 25, 2026. The article cites external research: University of California, Riverside (reported by The Washington Post) finding that every 100-word AI search uses roughly 519 milliliters of water, and a LinkedIn (2025) stat that 66% of leaders would not hire someone without AI literacy.
Details/background: Interviewees noted concrete environmental impacts — AI/data centers require large amounts of energy and water (cooling towers) which can impact nearby cities’ water pressure and quality and increase carbon emissions. They recommended using AI as a service, not a replacement, and urged cross-disciplinary action involving environmental experts, policymakers, researchers, and developers to pursue more sustainable AI practices.
-
Render Networks Announces $14.3 Million Funding Raise From Black Kite Partners
Render Networks announced $14.3 million in private equity growth funding and the acquisition of mPower Innovations.
- Acquisition and funding: Render Networks will acquire mPower Innovations and announced $14.3 million in private equity growth funding from BlackKite Partners; as part of the acquisition Render will transition mPower’s digital modeling platform to Esri’s ArcGIS. Greg Calcari (mPower founder) and Jason Brown (mPower CEO) will remain in senior leadership roles at Render. The company is based in Australia with U.S. headquarters in Denver. No specific implementation timeline for the transition or integration was provided in the announcement.
- Context and background: Render provides an AI-driven SaaS platform for construction data and cites use in BEAD deployment and major infrastructure projects; past customers named include Lumen, Connect2First, and Craighead Electric Cooperative Corporation. BlackKite Partners is an Australian private equity firm that launched in March 2026. The announcement includes executive statements from Stephen Rose (Render CEO) and Adrian Kerley (BlackKite Partners).
-
Almost 40% of data center projects will be late this year, 2027 looks no better
The Financial Times has published a study finding widespread delays across planned data center openings.
- Main finding: The FT analysis, using SynMax satellite imagery and permit data compiled by IIR Energy, found that almost half of data centers scheduled to open this year are likely to be at least three months late and more than 60% of projects scheduled for next year have yet to begin construction. The analysis cites project-level progress on land clearance and foundations and names major projects tied to Microsoft, Oracle and OpenAI.
- Background and causes: The report attributes delays to “chronic shortages of labor, power and equipment”, specialist trades shortages (electricians, pipe fitters), parts and component shortages (GPUs, memory, hard drives), permitting delays, and local/state pushback (for example Maine has paused large data center builds). OpenAI and Oracle issued statements saying their OpenAI-linked sites in Abilene, Shackelford County and Milam County, Texas are progressing on schedule, and OpenAI cited partnerships with Oracle and SB Energy for those projects.
-
The Genesis Mission: How AI Supercomputing Is About to Reshape American Science and Energy
The U.S. Department of Energy (DOE) has launched the Genesis Mission, chartered to double U.S. R&D productivity within a decade by deploying a platform combining high-performance computing, AI supercomputing, and quantum computing.
- Main action: The DOE’s Genesis Mission is standing up national AI supercomputing infrastructure through the Genesis Consortium with 27 industrial partners, including Nvidia, Oracle, AMD, and HPE; Argonne will host a system with ~10,000 GPUs (operational this year), Oak Ridge will host a comparably sized cluster targeting 2026, and a 100,000-GPU cluster is planned for Argonne in 2027. The program pairs this compute platform with a portfolio of national challenges (energy, physical sciences, national security) and a university engagement effort to train future scientists in AI-enabled methods.
- Background and concrete details: The initiative was launched by President Trump and chartered through the DOE; examples cited include fusion surrogate models that run thousands to tens of thousands times faster than traditional simulations, Grid FM from Brookhaven that could cut a ~20-year grid-simulation workload to two months, and DOE Office of Electricity efforts to reduce interconnection delays by addressing the 80–90% deficiency rate in interconnection applications. Named private partners and startups involved include Periodic Labs, Radical AI, and the Prometheus Project.
-
Cisco extends its Enterprise Agreement to include Nutanix Cloud Platform
Cisco has extended its Enterprise Agreement to include the Nutanix Cloud Platform, adding Nutanix HCI to Cisco’s EA licensing and services program.
- Main announcement: Cisco has officially extended its Enterprise Agreement (EA) to include Nutanix Cloud Platform (HCI), providing customers with predictable pricing, price protection for the EA term, and flexible, true-forward consumption (ability to increase Nutanix usage during the year and pay at the annual anniversary) without renegotiating contracts.
- Background and implementation details: Cisco and Nutanix have had a multi-year partnership (Cisco ended development of HyperFlex in 2023 and handed HCI to Nutanix); the vendors deliver products such as Cisco Compute Hyperconverged with Nutanix (combining Cisco hardware and Nutanix Cloud Platform), tightened Intersight–Nutanix integrations, support for Nutanix GPT-in-a-Box, and e-bonded global support and remote cluster deployment capabilities (per World Wide Technology commentary).
-
Firmus appoints three directors for AI infrastructure push
Firmus Technologies has appointed Lee Hatton, Christine Bartlett and Julie Shuttleworth as non-executive directors of Firmus Grid.
- Appointment details: The three appointees will take on committee and oversight roles; Lee Hatton will chair the Audit and Risk Committee, Christine Bartlett will chair the Remuneration and Nomination Committee, and Julie Shuttleworth joins as a non-executive director. The hires are presented as an announcement of new board appointments to support Firmus’s expansion of its AI infrastructure business with a focus on energy efficiency and remote/complex operating environments.
- Background and context: Hatton brings 20+ years across financial services, strategy and risk (previous board role at Xero, executive roles at NAB, Suncorp, Afterpay, Block); Bartlett has board/executive experience across property, IT and finance (chair of NSW Ports, CEDA, prior roles at IBM, National Australia Bank, Jones Lang LaSalle); Shuttleworth has 30+ years in mining, resources and energy including senior roles at Fortescue and Fortescue Future Industries. No monetary values or deal sizes were disclosed in the announcement.
-
TrendAI, Nvidia bring digital twin security to AI hubs
TrendAI has released an integration with NVIDIA’s DSX Air platform to let organisations model and assess security controls for AI-centric datacentres using digital twin simulations before physical deployment.
- Integration details and purpose: The collaboration links TrendAI security tooling with NVIDIA DSX Air digital twin simulations to validate network design, performance, and security for so-called AI factories during early-stage planning and before hardware install; key technical elements include TrendAI Vision One agentless EDR deployed on NVIDIA BlueField DPUs integrated with DOCA Argus, and TrendAI TippingPoint for network defence and virtual patching (drawing on TrendAI’s Zero Day Initiative and NIDS/NIPS capabilities). The integration supports red-team exercises using the MITRE framework to model adversary behaviours and evaluate control effectiveness.
- Background and scope: DSX Air is described as a cloud-hosted network simulation environment intended to shorten evaluation cycles and reduce reliance on physical labs; TrendAI cites IBM research noting more than 1 in 10 organisations experienced data breaches involving AI models/apps last year and that organisations without AI/automation faced USD $1.9 million higher average breach costs, and the partners position the integration for use in regulated sectors such as finance, healthcare, and government for evidence of control effectiveness and operational assurance.
-
Amazon leads concentrated global hyperscale market
The Business Research Company reported that Amazon led the global hyperscale data centre market in 2024.
- Main finding: The report states the top 10 providers captured 46% of global hyperscale data centre revenue in 2024, and notes Amazon held a 9% global market share in 2024 (the report’s regional breakdown later allocates 10% to Amazon); the top-10 list includes Amazon, Microsoft, Alphabet, Oracle, Meta, Apple, Alibaba, Equinix, Huawei and Tencent.
- Report details and technology priorities: The sector is shaped by rapid capacity expansion for cloud and AI workloads, investment in higher-density designs, liquid cooling, modular builds and energy efficiency, plus power procurement deals and partnerships; the report highlights AI-driven operations (“autonomous intelligence“, “self-healing cybersecurity frameworks“) and cites Adani Group’s plan to develop renewable-energy-powered, hyperscale, AI-ready data centres targeting 2035 as a “sovereign energy-compute platform“. Regional leader examples (North America, Asia Pacific, Western/Eastern Europe, South America) and institutional investors (e.g., Canada Pension Plan Investment Board, Crow Holdings Capital Partners) are listed.
-
Cisco grows high-end optical support for AI clusters
Cisco has announced the Open Transport 3000 Series multi-rail open line system and several accompanying optical products and upgrades.
- Main announcement: Cisco unveiled the Open Transport 3000 Series, a multi-rail open line system that integrates optical components for multiple fiber rails into a single line card to improve power, density, and capacity for hyperscalers, neocloud operators, and large enterprise AI; Cisco also announced the NCS 1014 (1RU, 800GE line card with 12.8T capacity and MACsec support) and a QSFP-DD Pluggable Protection Switch Module with sub-50 ms failover and ~90% rack space saving.
- Background and product details: Cisco stated AI optics TAM > $20B/year by 2030; NCS 1014 doubles density vs prior NCS generations, supports C&L-band and 800GE clients (map to a wavelength or inverse multiplex across two wavelengths); the QSFP28 100ZR 0 dBm coherent pluggable (Acacia-developed Bright) is targeted at edge/access/enterprise/campus deployments.
-
Cisco blends Splunk analytics, security with core data center management
Cisco has integrated Splunk IT Service Intelligence (ITSI) natively into its Nexus Dashboard to provide embedded analytics and observability for data center and campus networks.
- Integration details: Cisco integrated Splunk IT Service Intelligence (ITSI) with Nexus Dashboard, enabling collection and action on events, alarms, health scores, and inventory through open APIs; provides pre-built and customizable dashboards for inventory, health, fabric state, anomalies, and advisories; targets Nexus 9000 Series and Cisco 8000 Series switch/router environments.
- Background and specifics: Cisco acquired Splunk for $28 billion in March 2024; ITSI features include business and service monitoring, intelligent incident management, predictive analytics, and network topology monitoring; integration is described by Usha Andra and Anant Shah and is positioned to enable NetOps/SecOps convergence and streaming of high-fidelity telemetry to Splunk analytics.
-
Cisco: LPO not a panacea but plays strategic role in AI networks
Cisco has committed to developing Linear Pluggable Optics (LPO) options for its Silicon One family.
- Main announcement: Cisco will develop LPO options for the Silicon One family and has already shipped first commercially available 800G LPO modules for Nexus 9000 switches and 8000 series routers, targeting AI leaf-spine fabrics, data center interconnects, and high-density 800G deployments. The Silicon One G300 chip (102.4 Tbps, 512 lanes of 200 Gbps SerDes) is cited as enabling extended reach (up to 500 meters or 2 kilometers) and 30%–50% power reductions when Silicon One is deployed on both ends of the link.
- Background and implementation details: Cisco and its SVP Bill Gartner emphasize pairwise validation between LPO modules and host platforms (LPOs rely on host DSP/SerDes rather than onboard DSP), note reliability testing where 20 optics samples (100G/400G) failed Cisco stress tests, and say broader Silicon One-based systems and co-packaged optics (CPO) may follow with different trade-offs in scale and reliability. No monetary values or timelines beyond product introductions were provided.
-
How Environmentally-Conscious Businesses Can Help Save the Planet by Moving Toward Eco-Friendly Web Hosting
Artemis (Happy Eco News) advises businesses to adopt eco-friendly web hosting as part of their sustainability commitments.
- Main action: The article recommends switching hosting to providers that use direct renewable energy, renewable energy credits, or carbon offsets, and urges businesses to audit hosting energy disclosures and request documentation on energy sources and carbon programs before migrating. It cites concrete metrics: data centers used ~415 TWh in 2024, projected 945 TWh by 2030, and ~182 million tons CO2 in 2024 (International Energy Agency); migration practicalities include DNS propagation 24–48 hours and routine migration steps.
- Background and details: The piece references the EU’s Energy Efficiency Directive requiring annual reporting for data centers with installed IT power >500 kW and targeting an 11.7% reduction in energy use by 2030; it also cites consumer research: PwC found consumers willing to pay 9.7% more for sustainable goods, 85% affected by climate disruption, and Simon-Kucher found 64% rank sustainability among top 3 purchasing factors. The article is an advisory/guide rather than a novel policy announcement.