Microsoft will make Azure the first announced hyperscale cloud deployment of 3M's Expanded Beam Optical technology, but the partners have not disclosed its rollout, cost or measured advantage over existing fiber connectors.
Microsoft has attached the Azure name to 3M's Expanded Beam Optical technology, giving the connector its first publicly identified hyperscale cloud deployment. What the deal does not yet provide is the evidence needed to judge the purchase: rollout scope, price, measured labor savings or a comparison with the connectors Azure already uses.
Microsoft will deploy 3M's Expanded Beam Optical, or EBO, technology in Azure data centers. The companies described Azure Cloud and AI Infrastructure as the first announced hyperscale cloud provider to use it in their July 15 announcement.
That wording matters. It establishes the first hyperscaler willing to be named publicly, not the first commercial availability of EBO, an exclusive supply arrangement or an Azure-wide rollout.
The technical pitch concerns the physical interface between fibers. Traditional connectors transmit light through direct physical contact and require precise alignment. Contamination at that interface can trigger inspection and cleaning. 3M's design expands the beam across a non-contact interface, making a particle less disruptive to the optical path.
3M says this makes connections faster to install, more tolerant of contamination and easier to maintain. A technical account says the interconnect can be configured from 12 to 144 fibers for rack and infrastructure applications. Those configuration limits describe the product; they do not quantify its effect on an Azure build.
Microsoft said early use showed the potential to reduce network deployment timelines in certain environments and delivered strong signal performance in live data center conditions. The claim is deliberately narrow. Neither company disclosed the sites, number of connections, test duration, installation baseline or time saved.
The strongest evidence limiting the maintenance pitch comes from an industry project devoted to the same connector category. The project materials describe cleaning and inspection as a potentially substantial data center operating cost and expanded-beam connectors as less sensitive to contamination and scratches.
But the project also says expanded-beam end faces still require inspection and cleaning to prevent degradation of an optical link. It set out to estimate potential savings and compare expanded-beam connectors with physical-contact alternatives; the retained page does not publish the resulting cost comparison.
The supportable distinction is therefore reduced sensitivity and potentially less cleaning—not a maintenance-free connection or a proven Azure cost reduction. That boundary is particularly important because several accounts of the agreement repeat performance and maintenance claims that originate with the companies.
The Microsoft agreement is a commercialization milestone for 3M, but EBO was already a product. In a March manufacturing announcement, 3M said the technology had been commercially available since late 2024 and was in mass production.
3M also said a planned U.S. expansion would more than double capacity by adding manufacturing equipment and production space. It did not give the investment amount, current or future unit capacity, completion date or customer commitments supporting the expansion. The plan is evidence of supplier confidence, not proof that supply has already doubled.
A report citing 3M executives identified Microsoft as a previously unnamed hyperscale customer and said management had doubled its estimate of the addressable market to more than $2 billion. The report provides neither the underlying investor statement nor the assumptions behind that estimate, while the companies disclosed no financial terms for the Microsoft agreement.
The competitive-supply question is also unresolved. The Expanded Beam Optics Multi-Source Agreement group says it aims to standardize EBO connectors so multiple vendors can supply the industry. It listed 44 members as of July 7, including Microsoft, Meta, Nvidia, Cisco, Molex, Oracle, Senko, Sumitomo and TE Connectivity.
Membership shows interest across customers and suppliers. It does not yet demonstrate interoperability. At the research cutoff, the group's specifications section said released specifications would appear there but listed none. Until a specification and compatible products are available, the coalition cannot by itself answer whether buyers will have interchangeable sources or remain dependent on particular implementations.
The partnership is broader than an optical-component purchase. 3M will use Microsoft's AI and digital platforms in customer service, finance, sales and marketing as part of its enterprise transformation.
The most specific software project is a customer-order workflow. Engineers from Microsoft's Frontier Company initiative are working with 3M's Global Business Services team on an AI-agent system for credit checks, delinquency assessments and system updates. The companies say the design includes human approvals and a monitoring dashboard providing real-time visibility.
They expect the workflow to reduce manual effort, improve speed and consistency, and accelerate cash flow. Those are targets, not reported outcomes: the partners supplied no implementation date, cost, productivity baseline or forecast gain.
Each company is therefore adopting technology from the other, but the announcement does not value either side of the exchange. It also leaves unclear whether the two workstreams share commercial terms or are simply presented under one strategic-partnership label.
The next consequential step is not another partnership statement. It is evidence that EBO changes Azure's deployment economics without creating an unacceptable supply or compatibility tradeoff.
That judgment requires a small set of comparable results:
For now, Microsoft has given 3M a named hyperscale reference, and 3M has committed to expand manufacturing. The evidence does not yet show the size of the Azure deployment or whether its operating savings outweigh connector and supply-chain costs.
Zhipu reportedly reached $1 billion in annual recurring revenue in July, roughly four times a March estimate, but the unconfirmed run rate is not annual sales and still sits far ahead of recognized cloud revenue while margins remain thin.
Anthropic’s 20-year lease gives TeraWulf a customer for 401 megawatts of future AI infrastructure, but rent starts only after delivery and the proposed utility deal passes power and grid costs to TeraWulf, leaving financing and project margin unresolved.
OpenAI has put conversations and Projects back in its redesigned ChatGPT desktop app and enabled cloud Work threads to move across devices, correcting the launch's biggest usability failures without merging local Work or Codex histories.
Twenty-nine countries signed an agreement creating WAICO as an independent intergovernmental organization, while China paired the launch with capacity-building offers that are not yet confirmed as WAICO programs.
Meta reportedly plans to put departing AWS compute executive Dave Brown to work on its data-center buildout, adding hyperscale operating experience while leaving any customer-facing cloud business conditional and undefined.
Moonshot AI has made Kimi K3 available through its apps and API, pairing a 2.8-trillion-parameter architecture with early frontier-level results, but the model's open-weight claim cannot be tested until its weights and technical report arrive.
CIA Director John Ratcliffe said US intelligence is consistent with an estimate that Russian recruits last 20 to 30 minutes on Ukraine’s battlefield, but the public trail leads to an unsourced claim about assault troops and does not establish a representative average.
China has paired a five-year AI training offer for developing countries with cooperation centers, a weather-warning rollout and a new 29-country organization. The package gives Beijing a platform for influence, but no budget, selection rules or delivery timetable has been published.
Nebius has arranged its first senior secured facility against an operating GPU deployment and one customer’s cash flows. The deal adds project-level debt to its expansion toolkit, but an unnamed customer and a larger rival financing limit what it proves about the rest of Nebius’s backlog.
Huawei publicly displayed a 16-cabinet Atlas 950 configuration rated at 1 EFLOPS in FP8, providing tangible evidence of its system-scale AI strategy while leaving price, power use and sustained workload performance undisclosed ahead of the full system's planned fourth-quarter release.
UK testing places leading open-weight models four to seven months behind selected closed-model cyber results, yet longer attack chains, U.S. benchmarks and mixed cost comparisons show why that interval is a warning signal rather than a universal capability clock.
Moonshot AI's Kimi K3 added to a global technology selloff with near-frontier performance, but unreleased weights, mixed cost comparisons and a recommended 64-accelerator deployment leave its effect on chip demand unresolved.
Two binding EU decisions require Google to give rival AI services comparable access to 11 Android features and offer eligible search competitors a restricted, anonymized dataset, but phased deadlines, certification, pricing and privacy safeguards leave the competitive effect unproven.
Databricks has signed a term sheet for a Coatue-led financing at a $188 billion valuation, while unidentified sources put the round at $3 billion. The proposed capital would deepen its push into AI governance, data agents and operational databases, but the transaction remains open and the company supplied no new operating figures.
OpenAI says a handful of GPT-5.6 Sol file-deletion reports most commonly involved Full Access without sandboxing or Auto-review. Its own evaluations show a more complicated risk picture, and the company has not yet published an incident rate or evidence that its promised safeguards stop the failure.
Netflix says generative AI workflows were used on roughly 300 titles in 2026, but its only quantified example tied to that disclosure covers 17 minutes and does not establish how much finished material, spending or labor changed across the slate.
A rogue browser extension can trigger Claude for Chrome’s fixed Gmail, Docs, Calendar and business workflows; the attack is constrained by default approvals but can run silently for users who enabled unattended action.
Gemini 3.5 Pro missed its expected June rollout. An anonymous-source account says a late-June data change fell short of Google's coding goals, but Google has confirmed only partner testing—not the reported cause, a new date, or public results and pricing.
Gold Eagle has begun collecting and prioritizing AI-discovered software vulnerabilities, but the voluntary federal clearinghouse has not disclosed results, operating rules or the resources that would turn findings into deployed fixes.
Google is bringing Gemini Omni editing and a reusable face-and-voice avatar to Vids, but the sharper distinction is account-level identity across Vids and Gemini rather than a new category of AI video; Vids already offered Veo generation and customizable avatars, while specialist rivals already sell digital presenters.