TechCapsules

Microsoft reported a 27 % decline in Windows OEM revenue for the July‑September 2026 quarter, attributing the drop to the fading upgrade surge that followed Windows 10’s end‑of‑support deadline in October 2025, excess inventory at OEMs, and a weaker overall PC market. The revenue slump lowered the broader Windows OEM and Devices segment to $3.55 billion, down 22 % year‑over‑year. While Windows 11 now commands 71.34 % of Steam’s hardware survey share, Windows 10 still holds 23.61 % and grew 0.71 % in the same period, indicating a [...]

Sean Endicott, a seasoned Windows Central journalist, has reported on Microsoft’s software and hardware developments since 2017, covering Windows 11, Surface devices, Microsoft 365, and AI initiatives. He authors the weekly Windows Wrap column, curating key industry stories. Endicott earned a First Class degree in Broadcast Journalism from Nottingham Trent University and previously worked in live production. Outside tech, he coaches youth American football in Nottingham, earning the BAFCA Youth Coach of the Year award in 2024.

Eye Research uncovered that an abandoned Popcorn Time updater exposed up to 350,000 Macs to remote takeover, allowing attackers to run malicious code with root privileges. The vulnerability, CVE‑2026‑30612, stemmed from an unencrypted HTTP updater that accepted unsigned code. Between November 2025 and March 2026, researchers deployed a patch that disabled the updater on affected machines, distributing over two million updates before ceasing the operation in March 2026. The incident highlighted risks from legacy software infrastructure.

Mintive, an agent‑management startup founded by former Coursera engineer Jiquan Ngiam and TensorFlow co‑creator Vijay Vasudevan, secured $11 million in funding, including a $7 million seed round led by Cowboy Ventures. Investors also include Coatue, Operator Collective, Hustle Fund, and angels such as Jeff Dean and Andrej Karpathy. The company offers an MCP gateway, LLM gateway, and agent monitor to simplify enterprise AI agent deployment, claiming rapid, button‑click setup. Since launch, it has attracted over 50 enterprise customers, including [...]

Strava has extended its AI integration suite by linking its fitness platform with Google Gemini, enabling subscribers to receive personalized workout insights through natural‑language queries. The move follows a prior partnership with Anthropic’s Claude via the Model Context Protocol, which allowed users to analyze training data through AI conversations. The new Gemini integration is read‑only, optional, and initially limited to U.S. users, but it is available to all Strava subscribers regardless of their Gemini subscription level. The [...]

Google unveiled “Immersive Explorations” in its Gemini platform, a visual learning tool that lets students dissect subjects layer by layer. The feature, added to Gemini’s student hub, offers interactive, image‑based lessons on topics such as quasars, dinosaurs, plant molecular structure, and lightning physics. Users can click through visuals to access deeper details, with explanatory text accompanying each layer. The library is currently available in English, Indonesian, Hindi, Brazilian Portuguese, French, and simplified Chinese, with other languages to [...]

Microsoft engineer Raymond Chen explained why Windows 11’s Task Manager and taskbar can freeze when a CPU‑intensive application monopolizes processing time. When a heavy app consumes most of the CPU, the scheduler allocates limited time slices to all other threads, including Task Manager, which must wait for CPU cycles to render its UI and update the process list. Raising Task Manager’s priority would not help because a high‑priority thread can still be blocked on a lock or resource held [...]

Microsoft reported that its cloud segment generated over $63 billion in revenue for the fiscal year, marking a 30 percent year‑over‑year increase. The earnings call highlighted accelerated growth driven by AI‑focused services, including the Foundry platform, which saw a five‑fold rise in customers deploying multiple models. Executives emphasized a shift toward “intelligence per task,” aiming to improve cost efficiency and performance across heterogeneous AI workloads. The company also outlined plans to build a new technology stack designed for the [...]