Getting your news
Attempting to reconnect
Finding the latest in Climate
Hang in there while we load your news feed
Microsoft
Data center news, project activity, and monthly briefings for Microsoft.
Microsoft · Construction & power moves · 5
full tracker →Where Microsoft is securing land and power — each traced to primary filings.
Editor's picks
-
Assessing Sovereign AI: A Two-Pronged Framework
The article argues that sovereign AI is not a binary choice but a spectrum of partial sovereignty, and it proposes a two-pronged framework for assessing why states pursue it and how they build it across the AI stack.
- The piece compares five country cases — the United States, China, India, Singapore, and France — and maps their motivations across layers such as chips, compute, cloud, models, data, and applications.
- It highlights policy issues including interoperability, regulatory divergence, and the economic tradeoffs of building fragmented AI stacks; no single new project or deal is announced in the article, which is primarily an analytical commentary based on national documents and prior statements.
-
Microsoft to Lead Artificial Intelligence (AI) Masterclass at African Energy Week (AEW) 2026’s Renegade Intel Platform
Microsoft is announcing that it will lead an AI and Azure masterclass at African Energy Week 2026 in Cape Town.
- The Microsoft Masterclass: AI & Data Adoption for African Energy Companies will take place during the Renegade Intel platform at AEW 2026 in Cape Town from October 12–16 and will be led by Darryl Willis, Microsoft’s Corporate Vice President for Energy.
- The session will focus on AI, cloud computing, Azure, predictive analytics, IoT, cybersecurity, and digital transformation for Africa’s upstream, power, and renewable energy sectors; the article also references the City of Cape Town’s approval of two hyperscale data centers with a combined planned demand of 174 MW and Cassava Technologies’ Johannesburg Azure peering location.
-
AT&T and Microsoft scale trillion-token workloads with Microsoft Foundry and AMD
AT&T has announced that it used Microsoft Foundry Managed Compute to build and scale its OTel2.0 telecom-focused AI models.
- AT&T used approximately 530 GPUs through Microsoft Foundry Managed Compute, including 430 AMD Instinct™ MI300X GPUs, to support OTel2.0 development across multiple GPU architectures.
- The company processed about 1T tokens for OTel2.0 and says using open-source models like Phi-4 for synthetic data generation saved tens of millions of dollars versus frontier models; Microsoft also said deployment could happen in days rather than weeks.
- AT&T deployed models including Phi-4, OSS-120B, and Gemma-4 for synthetic data generation, data preparation, reasoning workloads, and broader model development.
-
NVIDIA Vera Rubin Driving Performance Per Watt, Lowest Token Cost for Partners Worldwide
NVIDIA has announced that Vera Rubin NVL72 is ramping into full production and is being deployed with partners CoreWeave, Google Cloud, Microsoft Azure and Oracle Cloud Infrastructure.
- Production ramp: Vera Rubin NVL72 racks are running at partners CoreWeave, Google Cloud, Microsoft Azure and Oracle Cloud Infrastructure, supported by a supply chain spanning 350-plus factory sites in 30 countries.
- Performance and infrastructure details: CoreWeave reported 10x more throughput per megawatt on DeepSeek-R1 versus Grace Blackwell NVL72; Google Cloud’s first A5X instance is up for London startup Ineffable Intelligence; Microsoft and Mistral are using Vera Rubin for a new multibillion-dollar agreement to expand AI infrastructure in Europe; the platform also features liquid cooling, chiller-free dry-cooler operation, and claims to save millions of gallons of water per megawatt annually.
-
Cassava Technologies strengthens Africa’s cloud resilience with Microsoft Azure ExpressRoute Metro designation in Johannesburg
Cassava Technologies has announced that Africa Data Centres was designated a Microsoft Azure ExpressRoute Metro peering location in Johannesburg, enabling a new Secure CloudConnect service through Liquid C2.
- Africa Data Centres became a Microsoft Azure ExpressRoute Metro peering location in Johannesburg, making it the first ExpressRoute Metro location in Africa and the second peering location within the Johannesburg metro.
- Liquid C2 will deliver Secure CloudConnect, a fully managed service combining private cloud connectivity and integrated cyber security; the release says the JHB1 facility supports the new local ExpressRoute Metro capability.
- The announcement quotes Ziaad Suleman and Vukani Mngxati and frames the move as supporting cloud resilience, business continuity, and AI-driven services across Africa.
-
Cassava Technologies strengthens Africa’s cloud resilience with Microsoft Azure ExpressRoute Metro designation in Johannesburg
Cassava Technologies has announced that Africa Data Centres has been designated a Microsoft Azure ExpressRoute Metro peering location in Johannesburg, and Liquid C2 will deliver the new Secure CloudConnect service.
- Africa Data Centres is now a Microsoft Azure ExpressRoute Metro peering location in Johannesburg, making Johannesburg the first ExpressRoute Metro location in Africa and giving JHB1 the second peering location in the metro.
- Liquid C2 will offer Secure CloudConnect, a fully managed service combining private cloud connectivity and integrated cyber security; the announcement cites growing regulatory focus on operational resilience, business continuity, risk management, and data protection.
- Ziaad Suleman said the designation is a “defining moment” for Africa Data Centres and Africa’s digital infrastructure, while Vukani Mngxati said the new capability gives organisations in South Africa a more resilient and secure path to the cloud.
Recent news
-
Reports: Google partnering with AMD for next-gen hybrid TPU
Google is reportedly working with AMD on the design of a next-generation TPU for AI and conventional computing workloads.
- The reported hybrid accelerator could integrate CPU cores directly into the TPU package, with a focus on reinforcement learning and agentic AI workloads.
- The article says Google already uses TPUs alongside standard x86 CPUs in servers, while AMD recently announced a partnership with Cerebras to combine processors for AI inference.
- This is a report about a rumored/industry-reported design effort, not a first-party Google or AMD announcement.
- Analyst Jon Peddie said the idea “makes good sense” because AMD has experience with custom APUs and Google “needs diversity.”
-
Data centers need power now. Natural gas is filling the gap
Bloomberg is reporting that rapid AI and data center growth is driving developers to build private natural gas power plants to avoid long grid connection delays.
- Ninety-nine proposed plants tracked by BloombergNEF could emit about 318 million metric tons of CO2 annually at industry-standard rates, or roughly 20% of current U.S. power-sector emissions.
- Many projects are concentrated in Texas; the story says some developers are considering solar power and battery storage, but natural gas remains attractive for around-the-clock power.
- This article is commentary/reporting on a Bloomberg investigation and references large projects involving Amazon and Microsoft, rather than announcing a new project itself.
-
Reimagining the ‘human side’ is key to unlocking tech-driven innovation in the Middle East
Strategy& argues that Middle East organizations should pair heavy tech investment with human-centric operating models to unlock sustained innovation.
- The article says innovation depends on empowered teams, flat decision-making, and talent models built for speed, accountability, and experimentation; it cites examples from Amazon, ING, GE, Google, and OpenAI.
- It references regional investment context, including $14.9 billion in Saudi tech and AI investment at LEAP 2025, a broader $100 billion AI initiative under Project Transcendence, and a $15.2 billion Microsoft-backed AI and cloud build-out in the UAE.
- This is an opinion/commentary piece, not a new corporate announcement; it summarizes known regional tech investment trends and recommends organizational changes rather than announcing a deal.
-
AI infrastructure needs clear environmental rules
Policy Circle published an opinion article arguing that India should set advance rules on power and water use for data-centre clusters supporting AI infrastructure.
- The piece says Microsoft, Google, Amazon Web Services and others have announced large investments in cloud and AI infrastructure in India, and cites growing demand in Hyderabad, Chennai, Mumbai, Bengaluru and Visakhapatnam.
- It argues for published environmental guardrails, cumulative reviews of grid and water demand, and standardized disclosure of electricity consumption and water use; it references experience in the United States, Ireland, the Netherlands and Germany.
-
Google, Microsoft and Nvidia back 800V DC standard for AI data centers
Google, Microsoft and Nvidia are working through the Open Compute Project to establish 800VDC as an open standard for high-density AI data centers.
- The companies are collaborating on common requirements for 800VDC power distribution, including power conversion, power quality, system interfaces, and safety.
- The article says the effort began being presented through OCP in 2025 and has moved toward formal specifications for the data-center supply chain.
- It cites potential efficiency benefits of 50% to 80% lower copper usage, 8% to 12% annual energy-related OpEx reduction, and $4 million to $8 million CapEx savings per 10 MW build.
- It also notes AWS has not committed to 800VDC and is reportedly redesigning AI data centers through its internal Titus initiative for GB200-class systems and future Vera Rubin hardware.
-
As AI power demand soars, Musk turns to natural gas
The Wall Street Journal reports that SpaceX is increasingly relying on natural gas to power its expanding AI infrastructure.
- SpaceX is building massive AI data centers and a $16.8 billion chip-manufacturing facility in Texas, and is installing its own natural-gas turbines and power plants because grid connections can take years.
- The report says SpaceX’s data-center power capacity could rise from about 2 GW to 10 GW or even 15 GW by end-2027; it also says SpaceX and xAI have used gas turbines in Tennessee and Mississippi, and that SpaceX has secured cloud agreements with Google and Anthropic.
-
Principal Firmware Engineer
Microsoft has posted a job opening for a Principal Firmware Engineer focused on backend firmware systems for AI accelerators such as MAIA.
- The role is based in Multiple Locations, India and requires 12+ years of experience, with responsibilities spanning architecture, debugging, validation, and productization.
- The posting lists required expertise in C/C++/Rust, multicore systems, and SoC bring-up, and says the engineer will serve as the primary technical liaison across hardware/software, validation, and integration teams.
-
IT infrastructure shortages are real and lasting. Here’s how to cope
Network World has published an analysis of IT infrastructure shortages affecting enterprise AI, networking, storage, and compute planning.
- The article cites analysts and vendors saying memory shortages are driving lead-time extensions to 6, 12, or 18 months, with memory prices up 50% to 200%, PC prices up 35% to 45%, and some server prices up over 125%.
- It also reports enterprise network equipment pricing is projected to rise over 20% in 2026 and by a further 3% to 5% into 2027, with no price reduction expected until end-2027; examples and advice are drawn from Gartner, TD SYNNEX, WWT, Backblaze, Cisco, and Equinix.
-
More Data Center Operators Commit to Abbott’s Texas Standards as Power Companies Endorse ERCOT Batch Framework
Texas data center operators and power companies have responded to Gov. Greg Abbott’s directive by publicly endorsing tighter standards for grid access, water use, and transparency.
- First coordinated corporate response: Skybox Datacenters, MARA, Digital Realty, QTS, Compass, Montera, and others said they will comply with Abbott’s standards and disclosure expectations tied to the ERCOT large-load queue review.
- Policy and project details: Abbott’s Aug. 3 order follows a June 10 directive and could lead to legislation on cost allocation, water-efficient cooling, annual reporting, and incentive changes; ERCOT’s Batch Zero groups projects of 75 MW and greater for a single study.
- Company examples: MARA said it curtailed about 550 MW during Winter Storm Fern; Hut 8 described 1.5 GW of Texas capacity and its 1-GW Beacon Point campus; NRG said its Texas BYOP project centers on a 1.2-GW plant and $3.2 billion investment.
-
US Big Tech’s AI race is becoming a race for monopoly
Policy Circle has published an opinion article arguing that Big Tech’s AI spending could entrench control over cloud computing, models, and digital markets as US regulation eases.
- The article says Amazon, Alphabet, Meta and Microsoft are spending heavily on AI infrastructure, with combined capital expenditure of about $170 billion in Q2 2026 and FactSet estimating roughly $1.5 trillion of combined spending on data centres and chips during 2026-2027.
- It adds that Alphabet spent $44.9 billion in the June quarter, Meta expects $130-145 billion in 2026, Microsoft added 31 data centres across five continents, and Amazon expects about $220 billion of capex this year; the piece is commentary rather than a new corporate announcement.
-
Agentic AI could force a rethink of enterprise AI server design, researchers say
Microsoft Azure and the University of Texas at Austin have published research arguing that agentic AI workloads need a new generation of workflow-aware AI servers.
- The paper says agentic requests are fragmented and heterogeneous, with workflows repeatedly crossing the CPU-GPU boundary; a CORAL-based controlled study found 580 LLM calls interleaved with 552 tool invocations.
- The researchers proposed Agora, a server architecture that reallocates CPU/GPU resources, separates scheduling and orchestration roles, and uses affinity-aware scheduling; in evaluation it lifted host CPU utilization by about 30%, increased generation throughput by 82%, and cut tail latency by 2.5x.
- The article is an explanatory analysis of a research paper and expert commentary, not a corporate product announcement or commercial deal.
-
As Data Centers Chase Solar Direct, India's Hyperscaler Pipeline Is Quietly Filling Up
The article argues that hyperscalers and data centers are increasingly driving renewable power contracts, using Google’s Oklahoma solar deal and related global transactions as examples, while highlighting a similar trend emerging in India.
- Google and RWE signed a 15-year PPA for the full output of the planned 155 MW Crooked Creek Solar project in McCurtain County, Oklahoma; construction starts later this year and commercial operation is targeted for 2028.
- The article also cites Vattenfall securing two Danish offshore wind projects totaling 1,800 MW and Sonnedix closing €730 million in financing for a 540 MW solar-plus-storage portfolio across France, Italy, Portugal and Spain; in India, it points to ACME Solar’s 25-year PPA with SECI for a 300 MW wind-solar hybrid project and says CleanMax’s data center client portfolio now accounts for 42% of its contracted capacity.
-
Land and Expand: The Gigawatt Credibility Test
This article is an analytical roundup of multiple recent U.S. data center announcements and one project termination, not a single new company release.
- The piece highlights new announcements from OpenAI, Microsoft, PowerPlay AI, Hut 8, Digital Realty, and Amazon, focusing on how projects are pairing scale with power supply, financing, water use, and community commitments.
- It also notes the termination of QTS’s portion of the Prince William Digital Gateway after litigation and local opposition, using it as a warning that entitlement and procedural issues can stop major projects.
-
New Data Center Developments: August 2026
This article is a curated monthly roundup of data center news, summarizing multiple announcements, reported deals, and policy actions across regions.
- It highlights several new project announcements and construction updates, including Meta’s 5 GW Hyperion expansion, OpenAI’s Project Camellia in Georgia, Iron Mountain’s 48 MW RCH-1 break ground, and Meta’s 1 GW Canadian campus in Alberta.
- It also covers policy and market developments such as New York’s permit pause for sites over 50 MW, Texas’s ERCOT queue audit, North Carolina’s tax change, Nebraska’s incentive block, plus large capital figures including €1.5 billion, $38.4 billion, $500 million, $400 million, and $10 billion across various projects.