Jul 14, 2026

Why Performance per Watt Is the Ultimate Metric for AI Infrastructure Efficiency
NVIDIA has published a blog post arguing that power efficiency is the defining constraint for AI factories and highlighting the rack-scale performance of its Blackwell and Vera Rubin platforms.
- The post says NVIDIA GB300 NVL72 delivers up to 25x performance per watt versus Hopper on leading open models, with additional examples of up to 20x on GLM5.1 and 10x on Kimi K2.6.
- It also says NVIDIA DSX MaxLPS can help operators run up to 40% more GPUs within the same power budget, and cites production use by Anthropic, OpenAI, CoreWeave, Perplexity, and Fireworks AI.
- This is a commentary-style NVIDIA blog, not a standalone product launch announcement, though it references existing deployments and platform capabilities.
Jul 9, 2026

Frontier models and production agents: Advancing Microsoft Foundry for the agentic era
Microsoft has announced new generally available capabilities in Microsoft Foundry, including GPT-5.6 models, the Asia-Pacific Data Zone, and hosted agents in Foundry Agent Service.
- GPT-5.6 Sol, Terra, and Luna are now generally available in Microsoft Foundry Models and Foundry Agent Service, with deployment options across Global Standard, Global Priority Processing, Data Zones Standard, and Global Provisioned.
- The Asia-Pacific Data Zone is also generally available, and hosted agents gain production features including VNet integration, Voice Live, Foundry IQ, toolboxes, and upcoming publishing to Microsoft Teams and Microsoft 365 Copilot next week.
- The post also says tracing and evaluation for hosted agents are generally available, while memory, routines, agent optimizer, and ROI for agents are in preview.
- This is a product announcement and platform update from Microsoft, not a report on a new external deal; it includes usage examples from Adobe, Telefónica, Tata Consultancy Services, and Viva Republica (Toss).
Jul 7, 2026

Financial Stability Report - July 2026
The Bank of England has published its July 2026 Financial Stability Report, warning that AI-related valuations, leverage, private credit risks and frontier AI cyber threats are increasing financial stability risks.
- The report says vulnerabilities in risky asset valuations, sovereign debt markets, private credit, and equity leverage remain, while frontier AI is creating new cyber and operational resilience risks.
- It also announces proposed reforms to the UK capital framework, including making buffers more releasable and consulting on changes to the leverage ratio; the report references earlier announcements and ongoing consultations rather than a single new deal.
- The report includes a system-wide exploratory scenario on private markets, states the UK has the largest data centre pipeline in Europe, and notes that AI data centre build-out will require significant investment and financing.
Jul 5, 2026

🎙️ Tranches de Tech #30 – Makers de père en fils
OVHcloud published a French podcast episode page featuring Sylvain Gougouzian and a roundup of tech links on AI, cloud, security, and conferences.
- The page presents the episode “Tranches de Tech – 30 makers de père en fils” with Sylvain Gougouzian, recorded on 30 June 2026, and links to the episode plus Sylvain’s social profiles and sites.
- The article also lists related items including Home Assistant x OVHcloud AI Endpoints, GitHub Copilot usage-based billing, NVIDIA NVCF, OpenAI acquires Ona, and OVHcloud posts on Terraform S3 state locking and secure image signing with Cosign and OVHcloud KMS.
- It is a curated news-and-links post / podcast description, not a standalone product launch announcement, and it references multiple external stories rather than making a single new commitment.
Jun 30, 2026

What the OCI MSA Didn't Solve for AI Scaling
Scintil Photonics’ CEO Matt Crowley argues that the architecture for optical scale-up has been settled by the OCI MSA and that manufacturing — specifically heterogeneous integration — now determines who can scale beyond four wavelengths per fiber.
- Main announcement/action: The article presents the argument that the Optical Compute Interconnect MSA (formed earlier this spring by AMD, Broadcom, Meta, Microsoft, NVIDIA, and OpenAI) settled on an NRZ modulation + wavelength multiplexing architecture (starting at four wavelengths per fiber), and that the remaining challenge to move to 8, 16+ wavelengths is industrial (manufacturing) rather than architectural. It identifies heterogeneous integration (bonding III-V gain material to silicon photonics wafers) as the manufacturing pattern that provides wavelength-scaling headroom and cites SHIP™ on Tower Semiconductor 200 mm lines and LEAF Light™ demonstrated in 8- and 16-wavelength configurations (with NVIDIA among Series B investors) as production proofs.
- Background and details: The piece contrasts prior eras (discrete lasers; silicon photonics with off‑wafer lasers) with the current heterogeneous integration era, notes that discrete-laser assembly scales poorly for hyperscale (e.g., a 16-wavelength source multiplies lasers and alignments across fibers), and references OFC 2026 where multiple vendors requested SHIP™ extensions across device categories. It emphasizes adding wavelength-scaling headroom as a line item on supplier evaluation sheets and states that teams delaying this consideration will need to redesign across two generations.
Jun 24, 2026 · Information Technology and Innovation Foundation

Economic Consequences of Section 232 Tariffs on Semiconductor Imports
The Trump administration issued an executive order in January 2026 imposing a 25 percent Section 232 tariff on certain advanced semiconductors.
- Main action: The January 14, 2026 executive order imposed an immediate 25 percent ad valorem duty on a narrow category of advanced computing chips and signaled a possible broader phase 2 at a “rate of duty that is significant.” ITIF models that a sustained 25% tariff would produce a cumulative $1.6 trillion loss in U.S. GDP (3.9%) over 10 years, reduce ICT consumption by $12.5 billion, and lower GDP per capita by $170 in year 1 and $4,825 cumulatively by year 10. The United States imported $48.1 billion of semiconductors in 2025 (baseline used in analysis).
- Background and recommendations: ITIF recommends removing blanket semiconductor tariffs, extending the investment tax credit (ITC) — now 35% — through 2030 and expanding it to semiconductor research and design, and requiring future Section 232 tariffs to include annual reviews and automatic expiration. The report also urges that tariff offsets under phase 2 be made available to firms engaged in semiconductor R&D and AI/data-center investment. The ITC is noted as scheduled to expire at the end of 2026.

Recent news

Jul 15, 2026

🤖 La Machine #85: RAISE Summit and AI’s Megawatt Moment
The article is a recap and commentary on the 3rd edition of RAISE Summit in Paris, highlighting a shift in AI discussion toward infrastructure, power, chips, data centers, and networks.
- 9,000 people attended the summit at the Carrousel du Louvre on July 8 and 9; President Emmanuel Macron opened with a video address arguing Europe needs its own AI infrastructure.
- The piece frames the event as a buildout and megawatts moment, with agenda topics including energy, chips, data centers, networking, storage, inference, and security.
- It also quotes Grant Lee, CEO of Gamma: “AI is the steel of our generation,” and notes the crowd was focused on whether enough infrastructure can be built to keep up with AI demand.
Jul 15, 2026

New York becomes the first US state to establish a moratorium on data centres
New York has announced a one-year moratorium on new hyperscale data centres while state regulators develop a new framework for the sector.
- Governor Kathy Hochul signed an executive order pausing State environmental permits for new hyperscale data centres for up to one year and directing agencies to assess electricity demand, water resources, and local community impacts.
- The order requires DPS to prepare a Generic Environmental Impact Statement (GEIS), Empire State Development to publish a Community Investment Framework within 60 days, and DEC to review water withdrawal rules within 12 months; it also sets a 50 MW threshold and proposes a New York Grid Acceleration Fund and a Data Center Interconnection Working Group.
Jul 14, 2026

Hitachi Energy to deliver 110kV grid connection for Kauri CAB data center in Frankfurt, Germany
Hitachi Energy has announced a contract to deliver a new grid connection solution for a Kauri CAB Digital Infrastructure data center in Frankfurt, Germany.
- Hitachi Energy will supply a dedicated 110kV grid connection designed for high reliability and a compact footprint to help integrate the data center into a constrained urban environment.
- The company said the solution will reinforce the interface between the transmission and distribution grids and support stable accommodation of growing electricity demand; no deployment timeline was provided. The article also says the project may relate to a 55,000 sqm parcel in Hochheim that Kauri acquired in April for €24 million ($28.2 million) and that the completed data center is expected to offer about 40MW.
- The piece additionally references prior Hitachi Energy deals: an April partnership with Samsung Engineering & Construction Group on HVAC grid infrastructure, a reported partnership with OpenAI on power transmission and distribution equipment, and a partnership with Southwest Power Pool on an Nvidia-powered AI tool for interconnection studies.
Jul 14, 2026

Yotta’s AI Infra Play, BUILT Bags $2 Mn & More
Yotta Data Services has announced a major AI infrastructure expansion and financing plan, alongside a potential public listing filing.
- Yotta plans to invest $7 Bn in AI infrastructure by FY27 and deploy 40,000 NVIDIA Blackwell GPUs over the next four months, taking total capacity to about 85,000 chips by FY27-end.
- The company recently raised $150 Mn at a $3.9 Bn valuation, partnered with US-listed Gorilla for GPU financing, and is preparing to file its DRHP for a $1.5 Bn public listing; the article also says most compute will be redirected to overseas markets.
Jul 13, 2026

White House plans new pledge to shield ratepayers from data center related bill hikes - report
The US government is expected to bring together data center firms and utility companies to announce a voluntary pledge aimed at preventing data center power demand from raising electricity costs for regular ratepayers.
- The pledge is expected to be announced at an event in the coming weeks; no company names were disclosed for the new pledge, though several major firms are expected to join.
- The story references earlier Ratepayer Protection Pledge signatories — Amazon, Google, Meta, Microsoft, OpenAI, Oracle, and xAI — and notes that states including Oregon, Oklahoma, Florida, Ohio, North Carolina, and Virginia have adopted or proposed rules making large-load data centers pay for new infrastructure costs.
Jul 11, 2026

OpenAI's Sam Altman accuses Musk of selling public market investors "on short-term space data centers"
OpenAI CEO Sam Altman criticized Elon Musk over SpaceXAI’s planned space-based data centers, accusing him of misleading public investors. - Altman said, “Homeboy you’re the one selling public market investors on short-term space data centers,” responding to Musk’s attack that Altman “takes scamming to a whole new level.”
- The article says Musk was a co-founder and early funder of OpenAI, later launched xAI, and that the combined company went public in June for more than two trillion dollars partly on promises of space data centers; it also says the company has promised up to one million compute satellites with the first AI1 design supporting up to 150kW peak power.
Jul 10, 2026

DCF Poll: How Much of the AI Data Center Pipeline Will Actually Get Built?
Data Center Frontier is posing a question about how much of today’s announced AI data center capacity will actually be operating by 2030.
- The article frames the issue as an industry poll/vote prompt, not a project announcement, asking readers to judge how much of the announced AI data center capacity will be built by 2030.
- It says future outcomes depend on securing land, power, permits, equipment, financing and community support, and notes that some projects may be delayed, redesigned, relocated or abandoned.
- The piece includes an editorial note that elements were created with help from OpenAI’s GPT5, and promotes Data Center Frontier’s social channels and newsletter.
- Contact information is provided for Matt Vincent at Data Center Frontier, including an email address and LinkedIn profile.
Jul 10, 2026

FLOPS vs Megawatts: Who’s Winning in 2026 Supercomputing?
The article provides analysis and commentary on 2026 supercomputing buildouts, contrasting public exascale systems with hyperscaler AI campuses. It is not a first-time announcement by one entity, but a roundup of recent developments and milestones.
- The piece compares public TOP500 systems and private hyperscaler AI campuses, highlighting that private builds are measured in hundreds of megawatts to gigawatts rather than HPL scores.
- It cites several recent milestones, including Microsoft’s Wisconsin Fairwater campus, xAI’s Colossus 2, OpenAI and Oracle’s Stargate, and Meta’s Prometheus nuclear power agreements.
- It also notes Alice Recoque installation in France under a €354.8 million EuroHPC JU contract with Eviden and mentions the next TOP500 list at SC26 in November.
Jul 10, 2026 · TechNode Global

Singapore Temasek bets big on AI: targets 10–15% AI allocation in portfolio by 2031
Temasek has announced a plan to lift its direct AI exposure from about 6% of its portfolio to 10-15% by March 2031, as outlined in its Temasek Review 2026 statement by CEO Dilhan Pillay.
- Direct AI exposure target: increase from about 6% to 10-15% by March 2031; current net portfolio value is S$518 billion ($402 billion) as at 31 March 2026.
- Temasek said the target implies AI net worth of roughly S$52 billion to S$78 billion, up from about S$31 billion now, and that the remaining 85%-90% of the portfolio should focus on AI adoption for competitiveness.
Jul 9, 2026

Singapore Temasek Reports Net Portfolio Value of $401 Billion (S$518 Billion) AUM in 2026 March, SGD Performance +10.5% in Last 12 Months, +7.1% 10-Year Return, +6.8% for 20-Year Return & +14% Since Inception in 1974, USD Performance +14.8% in Last 12 Months, +7.5% 10-Year Return, +8% for 20-Year Return & +15% Since Inception in 1974, Portfolio Allocation: 1) Temasek Portfolio Companies (TPCs) 43%, 2) Global Direct Investments (GDIs) 38%, 3) Partnerships, Funds & Asset Management Companies (PFAs) 19%, Portfolio 10-Year IRR – TPCs +8.1%, GDIs +7.6%, PFAs +7.7%
Temasek has announced its net portfolio value and performance for the financial year ended 31 March 2026, alongside a refreshed organizational structure and updated investment priorities.
- Net portfolio value rose to S$518 billion (US$401 billion) as at 31 March 2026; Temasek said it invested S$51 billion and divested S$31 billion, resulting in a net investment of S$20 billion.
- Temasek also said it has established Temasek Singapore (TSG), Temasek Global Investments (TGI), and Temasek Partnership Solutions (TPS) since 1 April 2026, and plans to increase AI-related exposure from 6% to up to 15% by 2031 while scaling core-plus infrastructure exposure from 1% to 5% by 2031.
Jul 8, 2026

Google-linked Housebound Group files for two data center projects in Haskell, Texas
Google-linked affiliate Housebound Group LLC has filed two new data center projects in Haskell, Texas, with TDLR.
- Journey 1A and Journey 2A are both planned for 1362 Red Creek Road, Haskell and each is budgeted at $400 million.
- Journey 1A began construction on April 1, 2026 and is expected to finish by late August 2027; Journey 2A began on June 23, 2026 and is targeted for late October completion. The site is described as roughly 300 acres, and the Haskell County tax abatement for the Journey site was approved in June 2025.
- The article also references related Google-linked development in Haskell County, including the Thelma campus and another campus developed by Ascendant Site Solutions LLC; Google said in November 2025 it would invest $40 billion in Texas through 2027.
Jul 8, 2026

SpaceXAI wants to compete on AI infrastructure, not just AI models
SpaceXAI has announced a new unified company strategy combining SpaceX and xAI to pursue orbital AI infrastructure.
- The company said it will combine Grok models, Colossus GPU clusters, Starlink networks, and SpaceX launch capabilities to build data centers in space powered by solar energy.
- It said it will deploy AI compute satellites as early as 2028 and begin work on the 11-million-square-foot Gigasat factory as soon as late 2027; the article also cites $55 billion for that factory investment.
- The article references prior disclosures that SpaceX spent $12.7 billion on AI in 2025 and that Anthropic and Google signed access deals worth $1.25 billion per month and $920 million per month respectively for Colossus.
Jul 7, 2026

Roundup: New Fiat EV / Microsoft using its own AI / Workforce housing
Microsoft has announced that it is replacing some OpenAI and Anthropic models with its own in-house MAI models in selected apps.
- Microsoft is already using its MAI models to handle tens of thousands of AI prompts each week in some apps, including Excel and Outlook, while routing more complex requests to external models when needed.
- Separately, Fiat Topolino is launching in the U.S. at $13,995 for use in private communities, resorts, and golf courses; Caddo Parish commissioners also halted a proposed temporary workforce housing ordinance tied to projects such as the Amazon data center.
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.

Tracking OpenAI's competitors too?

Book a 20-min call