Farewell to the 'Performance Black Box': Windows 11 Task Manager Officially Supports NPU Monitoring

marsbitPubblicato 2026-03-31Pubblicato ultima volta 2026-03-31

Introduzione

Microsoft has introduced a significant update to the Task Manager in a recent Windows 11 development build, adding real-time monitoring for Neural Processing Units (NPU). This enhancement addresses a key gap in hardware performance visibility, particularly as NPUs become central to AI PCs. Users can now track NPU utilization across processes, users, and details tabs in the latest preview version (26300.8142). The update also provides insights into dedicated and shared memory usage by applications. Additionally, the performance page displays detailed activity for neural engines integrated within GPUs, offering comprehensive oversight of all AI-related operations. This upgrade integrates NPUs into the mainstream hardware performance evaluation system, providing developers and power users with essential data for optimizing AI application efficiency and power consumption. The feature is currently available to Windows Insiders, with a broader rollout expected in the future, signaling deeper integration of AI hardware into the Windows ecosystem.

Microsoft recently introduced a major upgrade to the Task Manager in a Windows 11 development build update. This update finally completes the last piece of the hardware monitoring puzzle by officially adding real-time monitoring capabilities for the Neural Processing Unit (NPU).

With the popularization of the AI PC concept, the NPU has become the core hardware for handling artificial intelligence tasks. Previously, users were unable to visually check the operating status of the NPU, but this upgrade makes the load conditions of this dedicated chip transparent and visible.

In the latest preview build 26300.8142, the Task Manager's Processes, Users, and Details pages have all added an NPU column. Users can not only view the real-time NPU utilization but also gain in-depth insights into the specific consumption of dedicated memory and shared memory by various applications.

Furthermore, Microsoft has refined the monitoring dimensions; the Performance page can now fully showcase the neural network engine activity built into the graphics card. This means that all AI-related operational traces within the system will be comprehensively captured by the Task Manager.

This evolution of the Task Manager marks the NPU's official entry into the mainstream hardware performance evaluation system. For developers and tech-savvy users, this is not just an increase in data but also a foundation for precisely tuning the energy consumption and efficiency of AI applications.

Currently, this feature is primarily available to Windows Insiders in preview builds, and a wider rollout will take some time. It is foreseeable that with the普及 of NPU monitoring, future Windows systems will undergo a deeper experience reconstruction centered around the AI hardware ecosystem.

Domande pertinenti

QWhat major update did Microsoft recently introduce to the Task Manager in Windows 11?

AMicrosoft introduced real-time monitoring for the Neural Processing Unit (NPU) in the Task Manager.

QWhy is the NPU monitoring feature significant for AI PC users?

AIt allows users to visually check the workload and status of the NPU, which is crucial for handling AI tasks, making the specialized chip's operations transparent.

QWhich Windows 11 preview build version includes the NPU monitoring capability?

AThe feature is available in the preview build version 26300.8142.

QWhat specific details can users monitor regarding NPU usage in the Task Manager?

AUsers can view real-time NPU utilization and see how programs consume dedicated and shared memory.

QWho currently has access to the NPU monitoring feature in Windows 11?

AThe feature is currently available to Windows Insiders in the preview version, with a wider rollout expected later.

Letture associate

Rhythm X Zhihu Co-host Web4.0 Theme Event: When AI Agent Takes Over On-Chain Permissions

Most discussions about Web 4.0 miss the point. The real question is not whether it is a marketing trend, but rather: who is gaining control over the underlying permissions of the internet? Historically, each iteration of the web has involved a transfer of authority downward: Web 1.0 was read-only; Web 2.0 allowed users to write but platforms owned the data; Web 3.0 enabled true ownership through on-chain assets and private keys. Web 4.0 continues this trend, but the transfer is not to users—it is to AI Agents. The current infrastructure is human-centric, designed around human limitations like attention span and memory. But AI Agents don’t need intuitive UIs, password resets, or sleep. This creates a core tension: an internet built for humans is now being used by entities without human constraints. Two key shifts are underway: the decline of traditional front-end interfaces (replaced by API-driven machine communication) and the replacement of human-centric identity systems (like passwords) with granular, on-chain permissions. A critical enabler is crypto infrastructure. AI can make rapid decisions but lacks independent payment channels and asset sovereignty. Crypto fills this gap. Platforms like Hyperliquid offer 24/7 markets, ideal for non-stop Agent operation. When Agents control wallets and private keys, they can both decide and execute—forming complete economic entities. The real narrative of Crypto × AI isn’t just buzzword synergy—it’s the convergence of complementary infrastructures. The deeper shift is not which products will succeed, but how the rules of economic systems will change when AI becomes a primary on-chain participant, operating at scale and speed beyond human capability.

marsbit10 min fa

Rhythm X Zhihu Co-host Web4.0 Theme Event: When AI Agent Takes Over On-Chain Permissions

marsbit10 min fa

Claude's 18 Digital Pets Are Here: Tamagotchi in Your Terminal

Claude, known for its powerful AI models, recently had 500,000 lines of code leaked, revealing several unreleased features. Among them, the most talked-about is an electronic pet named Buddy. Released on April Fool’s Day, Buddy is an ASCII-based virtual companion that lives in the Claude Code terminal. It has its own species, rarity, five attributes (DEBUGGING, PATIENCE, CHAOS, WISDOM, SNARK), and personality. Users can interact with it by petting, calling its name, or asking it to show its info card. The system uses a two-layer architecture: "Bones" (deterministic) and "Soul" (non-deterministic). A seeded random generator assigns each user a unique pet based on their user ID, with no option to reroll. There are 18 species with five rarity tiers, including a 1% chance for a legendary or shiny variant. Buddy reacts to the user’s coding behavior, offering comments in speech bubbles. It is designed to increase user attachment during long coding sessions. Despite its playful nature, the feature is seriously engineered and is part of Claude’s planned product roadmap, signaling a shift in AI tool competition from raw capability to user engagement.

marsbit16 min fa

Claude's 18 Digital Pets Are Here: Tamagotchi in Your Terminal

marsbit16 min fa

Bitcoin fees fall to 14-year low – Why BTC price remains range-bound

Bitcoin's market is experiencing a period of low activity and weakening demand, as indicated by a significant drop in transaction fees to a 14-year low of 2.5 BTC per day. This decline reflects reduced on-chain activity, slowing speculative flows, and a lack of urgency among users. Concurrently, Bitcoin ETF flows have turned negative, with consistent outflows in late March, signaling fading institutional interest and profit-taking. Despite holding support around $68,400 and showing some signs of buyer absorption, BTC's price remains range-bound below $71,300 due to a lack of strong conviction. The market appears to be in a fragile balance, awaiting the return of meaningful demand and spot inflows to drive a potential breakout.

ambcrypto17 min fa

Bitcoin fees fall to 14-year low – Why BTC price remains range-bound

ambcrypto17 min fa

Embodied Intelligence Breakthrough: Amap Fully Open-Sources Universal Robot Base Model ABot-M0

Embodied Intelligence Breakthrough: AutoNavi Open-Sources Universal Robot Base Model ABot-M0 AutoNavi has announced the full open-source release of ABot-M0, the world's first unified architecture-based embodied manipulation base model. This model is designed to enable "one general brain to adapt to multiple forms of robots," aiming to break down barriers between heterogeneous hardware and accelerate the adoption of embodied intelligence in industrial and household settings. ABot-M0 demonstrated exceptional performance in industry tests, achieving a task success rate of 80.5% on the Libero-Plus benchmark—a nearly 30% improvement over the previous benchmark, Pi0. It also set new state-of-the-art records on benchmarks like Libero and RoboCasa. The open-source release addresses long-standing challenges in the field, such as data isolation and deployment difficulties, by providing resources across three key dimensions: - **Data:** The UniACT dataset, the largest of its kind, with over 6 million real operation trajectories and full data pipeline tools. - **Algorithm:** The model architecture and training framework, featuring innovative components like Action Manifold Learning (AML) and a dual-stream perception architecture. - **Model:** End-to-end pre-trained models and a complete toolchain for out-of-the-box deployment, significantly lowering the barrier to adaptation. According to AutoNavi's ABot-M0 technical lead, this open-source initiative aims to build a bridge between academic research and industrial application, enabling robots of various forms to possess a smart, reliable, and universal "brain."

marsbit1 h fa

Embodied Intelligence Breakthrough: Amap Fully Open-Sources Universal Robot Base Model ABot-M0

marsbit1 h fa

Will Middle Management Be Replaced by AI? What Will the Future Company Structure Look Like

The article explores whether AI will eliminate middle management and reshape future corporate structures. It traces the historical evolution of organizations—from Roman military units to modern corporations—showing how hierarchical systems emerged to manage information flow under the constraint of limited "span of control." Middle management, matrix structures, and bureaucratic systems were all solutions to coordination challenges in information-scarce environments. AI, however, challenges this foundational premise. By enabling real-time modeling, understanding, and distribution of information, AI could replace human-centric coordination mechanisms. Examples like the AI firm "Moon Dark Side" illustrate radical experiments: no departments, titles, or traditional KPIs, with co-founders directly managing large teams and AI agents handling tasks from data processing to code generation. Block (founded by Jack Dorsey) is presented as a case study in building an "intelligent company." This model relies on two core components: a "company world model" (a real-time understanding of internal operations via digital traces) and a "customer world model" (built from real behavioral data, especially financial transactions). An intelligence layer uses these models to dynamically combine capabilities (e.g., payments, lending) to serve customers proactively, without pre-defined product roadmaps. In this structure, traditional roles shift. Middle managers are replaced by a system that handles coordination, while humans focus on individual contributions (ICs), direct responsibility (DRIs), or player-coach roles. The organization becomes flatter, faster, and more adaptive. The article concludes that AI is not just a tool for efficiency but a transformative force that could redefine organizational design, moving companies from human-led hierarchies to system-driven intelligence.

marsbit1 h fa

Will Middle Management Be Replaced by AI? What Will the Future Company Structure Look Like

marsbit1 h fa

Trading

Spot

Futures