LectureRechercheRadarCadre d'investissement
Connexion / Inscription
ToutTechnologieProduitsInvestissementBourseAutres

Lecture

2026-11-10115 publications

Actualiser

Chinese Open-Source OS openEuler Previews Agentic Infra Optimizations, Claims ~84% KV Cache Latency Drop in Ad Business

Sujet · openEulerPrise rapide
2026-11-10 14:00 GMT+8

According to a report by Zhidx on November 10, Liu Fei, Executive Director of the OpenAtom openEuler Community, previewed technical breakthroughs for Agentic Infrastructure (Agentic Infra) at a media pre-briefing for the Agentic Infra Operating System Conference & openEuler Summit 2026 on October 9. The announcements cover Agent observability and security governance, ultra-fast sandboxes, agent memory systems, and KV Cache optimization.

Liu stated that a single agent session might launch dozens or even hundreds of tasks simultaneously, and traditional sandbox startup times consume significant resources. openEuler plans to optimize from the operating system level—specifically file systems, memory management, and security mechanisms—to enable faster startup and higher-density concurrent sandbox execution.

Regarding key performance metrics, Liu claimed that in an advertising business practice, openEuler reduced cross-node communication by optimizing data transmission between DDR (Double Data Rate) memory and accelerator cards, resulting in a nearly 84% drop in KV Cache read/write latency and improved real-time bidding capabilities. This figure is self-reported by the community, with no disclosed test benchmarks or independent verification reports yet available.

openEuler released an operating system version for supernodes last year and has listed support for Agentic Infra as a key evolution direction. As AI applications shift from single-model calls to multi-step task execution, the role of operating systems in task scheduling, runtime isolation, and data management becomes increasingly prominent.

Sources :zhidx.com

Recherche

Dyson Embeds the Logic of "Confirming Cleanliness" Rather Than "Completing a Whole-House Run" into the Design of Its Robot Vacuum and Mop

Sujet · 戴森扫地机器人设计逻辑Matériel
2026-11-10 17:30 GMT+8

For over a decade, the robot vacuum and mop industry has developed a highly standardized operational logic: mapping, route planning, traversing rooms, and cleaning at set frequencies. This logic treats "covering the floor" as the task itself, but coverage is merely a means; making the floor clean is the true objective. When faced with spilled vegetable juice or sugary drinks, methodically completing a single run may not truly remove the stains and could even spread the dirt around.

At the Dyson Unveiled global launch event, Dyson introduced the R3 Nurovi Spot+Scrub UV Ai smart robot vacuum and mop, attempting to bypass this contradiction from the design source. After practical testing, Ifanr editor Peng Haixing pointed out that the core change in this machine lies not in increased suction power or wider roller brushes, but in setting "ensuring the floor is clean" as the top priority, rather than "completing a cleaning cycle."

What makes this logic possible is a closed loop from perception to verification. The machine retains the green light dust detection technology first used in Dyson's 2021 V15 Detect vacuum cleaner, exposing micro-dust difficult for the naked eye to see through low-angle lighting. On this basis, the newly added UV stain detection system utilizes fluorescence effects to reveal semi-transparent residues left by dried liquids such as cola, milk, vegetable juice, and pet excrement. The high-definition camera on the body continuously observes the floor, sending information to the AI detection system—which reportedly can identify nearly 200 types of stains and objects in home environments—to make bidirectional decisions: planning paths to avoid obstacles like toys and slippers, and adjusting cleaning strategies based on content and floor material when encountering stains.

What truly changes the working method is the feedback mechanism Dyson calls the "cleanliness algorithm." Upon encountering a stain, the machine performs targeted scrubbing, then uses the "dual-light system" combined with the camera to check the cleaning result; if the stain remains, it continues scrubbing until the recognition system determines it can move on. This means the robot's work is no longer limited by a fixed number of cleaning passes, but adjusts actions according to actual conditions, much like humans doing chores.

On the mechanical level, this logic is supported. The side brush's operation shifts from conventional "high-speed beating" to "inward guiding," with bristle angles and rotation speeds strictly limited to ensure debris at walls and dead corners is steadily swept into the main brush's negative pressure airflow zone. When dealing with hair tangles, soft rubber structures resembling spiral shells and petals are added to the base of the side brush; tangled hair loses its "fulcrum" and slips off, being sucked directly into the dustbin by negative pressure airflow without relying on cutting blades. For mopping, the 27-centimeter-wide living-water roller integrates a 12-point hot water spray system inside, combining high-frequency hot water wetting with scraper dirt removal so that every rotation contacts the floor with a clean roller, avoiding "using one dirty cloth to wipe the entire house."

Sources :爱范儿

Recherche

Bluetooth SIG Adopts HDT Standard, Boosting BLE Transfer Speed by ~4x to 7.5 Mbps

Sujet · 蓝牙HDT标准Matériel
2026-11-11 01:30 GMT+8

The Bluetooth Special Interest Group (Bluetooth SIG) announced on November 4, 2026, the adoption of a new standard named "Bluetooth HDT (High Data Throughput)," which increases the maximum data transfer rate of Bluetooth Low Energy (BLE) to 7.5 megabits per second (Mbps), approximately four times the traditional standard.

This performance leap is achieved by introducing a new physical layer modulation scheme with enhanced robustness. Beyond speed, HDT improves spectral efficiency to expand communication capacity, while shorter transmission times contribute to better energy efficiency and increased reliability and resilience.

According to the roadmap, the first application based on Bluetooth HDT will launch in 2027, featuring a new audio codec called "HRC" that supports lossless audio transmission at 96 kHz sampling rate and 24-bit depth. Detailed specifications are expected to be revealed at CES 2027 in Las Vegas. The standard is also positioned as key infrastructure support for future applications in extended reality (XR), AI devices, commercial and industrial solutions, and wearables.

Sources :EE Times Japan

Recherche

Red Hat Releases OpenShift 5: Agentic Troubleshooting and In-Place Upgrades Debut, RHEL Adds Post-Quantum Cryptography

Sujet · OpenShift 5Matériel
2026-11-09 23:30 GMT+8

US enterprise software company Red Hat announced on November 9, 2026, at KubeCon + CloudNativeCon North America in Salt Lake City, a major update to its Kubernetes-based application development platform, OpenShift 5. This marks the first major version upgrade in seven years, aiming to simplify hybrid cloud infrastructure management through artificial intelligence.

The core change in OpenShift 5 is the expansion of the generative AI assistant Lightspeed into an autonomous 'agent' capable of investigating issues. Administrators can now use natural language to generate system performance visualizations, such as querying CPU utilization over the past week overlaid with memory consumption and application errors. More significantly, the new platform introduces agentic troubleshooting: when cluster events occur, the system automatically examines related information, reasons through potential causes, and proposes remediation steps, while retaining final execution control with the administrator.

On the operational front, OpenShift 5 addresses long-standing upgrade pain points. Mike Barrett, Vice President and General Manager of Hybrid Cloud Platforms at Red Hat, stated that the update supports in-place upgrades from OpenShift 4, eliminating the need to reinstall software or migrate workloads. This change significantly reduces complexity for enterprises running long-lived applications and virtualized infrastructure. Additionally, the platform includes an AI-driven upgrade planner that identifies application dependency and API compatibility risks before updates begin.

In terms of security, OpenShift 5 enhances multitenancy isolation, network isolation, and automated certificate management, specifically introducing post-quantum cryptography for control-plane communications aimed at sovereign cloud providers and regulated organizations. Red Hat also announced upcoming versions of Red Hat Enterprise Linux (RHEL) 10.3 and 9.9, which similarly strengthen anti-tampering protection, AI-assisted system administration, and post-quantum cryptography support.

Chris Wright, Chief Technology Officer at Red Hat, noted that this update is a key step in infusing intelligence into the platform, enabling agent-based troubleshooting through simple interfaces and natural language interaction. As enterprises focus more on GPU resource utilization, OpenShift 5 also supports automated GPU resource allocation across AI inference and traditional enterprise workloads to improve the efficiency of expensive computing infrastructure.

Sources :SiliconANGLE

Recherche

NVIDIA Partners with Rebellions and SK Telecom to Integrate Custom NPUs into MGX Rack Systems

Matériel
2026-11-10 16:47 GMT+8

NVIDIA, Korean AI chip developer Rebellions, and SK Telecom announced a partnership on November 10, 2026, to develop sovereign AI infrastructure that combines locally designed processors with NVIDIA’s NVLink Fusion interconnect technology. Under the agreement, Rebellions’ next-generation R200 neural processing units (NPUs) will be integrated into NVIDIA’s MGX rack architecture, with deployment planned at SK Telecom’s AI data centers.

The core of this collaboration lies in achieving heterogeneous computing interconnection through NVLink Fusion. Rebellions plans to use this technology to connect its R200 NPUs via high-bandwidth, low-latency networking, targeting energy-efficient inference for frontier AI models, including multitrillion-parameter mixture-of-experts architectures. NVIDIA’s MGX rack architecture will provide standardized system designs, while its ecosystem of original design manufacturers and component suppliers could help Rebellions accelerate development and reduce deployment complexity.

The companies also plan to explore disaggregated inference, splitting AI workloads between different processor architectures. Under the proposed approach, NVIDIA GPUs would handle the compute-intensive prefill stage, while Rebellions NPUs would manage token generation during decoding. This division of labor is intended to improve inference throughput and power efficiency, potentially addressing two major challenges facing AI data center operators.

SK Telecom plans to deploy these systems within its AI data center infrastructure, connecting the hardware with proprietary foundation models, including A.X K2, and public-facing AI services. The initiative supports Korea’s government-backed “AI for All” project and aims to strengthen domestic control over AI infrastructure, models, and services. Korea’s Ministry of Science and ICT is supporting the collaboration as part of a broader effort to strengthen the country’s sovereign AI capabilities and domestic semiconductor industry.

Rebellions expects its R200-based systems to be ready for deployment in 2028, giving the partners time to develop and validate the rack-scale architecture before commercial rollout. “Today’s announcement marks a turning point, opening a new chapter for sovereign AI by connecting Korea’s AI silicon and models with global systems technology,” said Sunghyun Park, cofounder and CEO of Rebellions.

Sources :eeNews Europe

Recherche

Intel Explains AI Deployment Cost Strategy: Agentic AI Increases CPU Demand

Matériel
2026-11-10 16:41 GMT+8

Semiconductor competition is not just about whose chips are faster, but also about who can best configure cost, power consumption, and computing capability in the right places. At a next-generation processor architecture briefing held on November 10, Intel discussed everything from the Core 3 Series to the data center GPU D series and the next-generation Xeon Diamond Rapids. Underlying these discussions is a common question: as AI gradually enters daily work, how much compute power do users actually need, and how much are they willing to pay?

The strategy for the Core 3 Series (codenamed Wildcat Lake) is one of subtraction. It utilizes Intel's 18A process technology, with a maximum configuration of two performance cores and four low-power efficiency cores, paired with the Xe3 graphics architecture and a neural processing unit, bringing next-generation cores into mainstream price points. However, the costs of new process technologies do not disappear into thin air. Yu Xiaolun, Product Manager for Intel's Client Computing and Physical AI Group, stated that mainstream-priced products do not necessarily have to rely on previous-generation processes. To address this, Intel has adopted single-channel memory, streamlined input/output configurations, and uses organic substrate multi-chip packaging via UCIe chiplet interconnects to reduce packaging and system costs. These adjustments reduce the compute die area by approximately 38% and allow motherboards to be reduced from eight layers to six. This means that price competition for mainstream products must consider memory, circuit boards, and components together, rather than looking solely at the processor's selling price.

Regarding agentic AI, Zheng Zhicheng, Director of Commercial Business for Intel's Business and Marketing Group, presented an argument: the fact that CPUs are more important does not mean GPUs are unimportant. His logic is that agentic AI requires not only model inference but also general-purpose computing environments to execute tools, manage data, and coordinate tasks; therefore, CPU demand will also increase. This provides a new perspective for understanding server requirements, but the article notes that the usage volume of agent services cannot be directly converted into equivalent physical core procurement. Actual demand is still influenced by the number of simultaneous users, resource scheduling, and task execution time.

The GPU D series (codenamed Crescent Island) bets on memory capacity and deployment costs. It adopts the Xe3P architecture, focusing on inference applications, with announced specifications including 350 watts of power consumption, 1.5 TB per second of memory bandwidth, and 1.3 PFLOPS of FP8 computing performance. Intel-branded cards are configured with 160 GB of LPDDR5X, while partner designs can reach up to 480 GB. Large memory serves not only to hold model weights but also key-value caches required for long contexts and working data for multiple simultaneous users. Paired with standard PCIe and air-cooling designs, Intel hopes to lower the threshold for adoption in existing server rooms. In a Q&A session, Zheng Zhicheng clarified that 160 GB is not the capacity floor for the entire series; smaller capacity versions may exist, but the 128 GB and 96 GB examples shown on-site cannot yet be considered official specifications.

Sources :INSIDE

Recherche

PolarDB-X Team Practice: The Bottleneck for AI Agents Fixing Bugs Is Not Coding, But Ensuring Analysis Withstands Verification Through Actual Execution

Sujet · PolarDB-XMatériel
Vérifié 2026-11-10 16:30 GMT+8

At the beginning of the year, Sheng Taotao of the PolarDB-X team fixed a bug that left a deep impression on him. In the end, only two lines of code actually needed changing, but checking logs, reviewing monitoring data, and analyzing the issue beforehand took him an entire day. To reproduce the issue, he had to squeeze in repeated debugging sessions between meetings, and it took nearly a week in total to finish. After that, he still had to add test cases and follow up on code review until the code was merged. AI can already help write those two lines of code, but what truly consumes time is the work preceding it. Can we hand over that work as well to an Agent?

The team divided development tasks into two categories: new features and iterative optimizations require developers to lead solution design with Agent-assisted coding and testing; whereas handling alert tickets and fixing defects involves relatively well-defined problems, small code changes, and results that can be validated through testing. They attempted to let the Agent autonomously complete the latter category of tasks, with developers responsible only for reviewing the deliverables.

However, for the Agent to autonomously complete this closed loop, its output must first be ensured to be reliable. When discussing reliability, people often focus on the coding stage, hoping to guarantee code correctness through specific skills or constraints. Sheng Taotao pointed out that in defect repair, reliable reproduction is the most important step. Without reproduction, it is difficult to confirm whether the analysis is correct; if the analysis itself is wrong, even perfectly structured code may fail to solve the real problem. Based on this, they established three key requirements: incident information must be complete, the issue must be reliably reproducible, and the fix must be validated through testing.

Around these three requirements, the team undertook extensive tool and environment construction. For information retrieval, they encapsulated common platform functions such as monitoring, logs, and instance topology into CLI tools, allowing the Agent to autonomously query structured data rather than relying on screenshots. For platforms that are poorly maintained and lack ready-made Agent-friendly interfaces, they referenced the implementation and authentication methods of existing CLIs, having the Agent analyze underlying API requests from web pages and wrap them itself. Within about a week, they completed the integration of commonly used functions.

What truly made the team realize the severity of the problem was a low-probability concurrency bug. This bug had existed in the production environment for a long time, but because it occurred rarely and was difficult to reproduce, its cause could never be located. After receiving the incident information, the Agent produced an analysis report within minutes, detailing code locations, execution flows, and possible failure points; it looked professional and convincing. However, after spending about two hours validating it, the team found the explanation did not hold up. After feeding the results back to the Agent, it quickly produced a new analysis, which was also incorrect upon validation. After several rounds back and forth, a pile of reports accumulated, yet the problem remained unexplained.

Sources :Alibaba Cloud Community

Recherche

Google Rolls Out 'Skills' Across Workspace to Standardize Team AI Workflows

Matériel
Vérifié 2026-11-10 16:11 GMT+8

On November 9, 2026, U.S. tech giant Google announced that 'Skills' are now rolling out across Google Workspace Studio and core Workspace applications. This feature aims to help enterprises standardize and scale their AI usage by enabling reusable custom instructions that guide Gemini to apply specific team rules, templates, and reference files.

Users can create skills from scratch in Google Workspace Studio, start with turnkey templates, or edit Markdown-native (SKILL.md) files directly. Additionally, skills can be drafted and collaborated on in Google Docs, then shared via Studio so teammates can customize and enable them in their own workflows. Users can also prompt Gemini directly within Gmail, Docs, Slides, Drive, and Chat to create, refine, and test skills without leaving the app they are working in.

Once a skill is turned on, Gemini in these apps can automatically recognize when it is relevant and apply it, or users can explicitly invoke one or more skills using an @mention. Skills can also trigger Workspace Studio flows configured with 'Start manually' or 'On a schedule' starter steps. For example, marketing or sales teams can maintain approved brand voice and structure across email drafts, proposals, and pitch decks, while finance or legal specialists can standardize recurring contract reviews against guidelines stored in Google Drive.

Currently, creating, editing, testing, and sharing skills are available on the web, while invoking skills is supported on the web and in the Google Chat mobile app on Android and iOS. Notably, Skills in Workspace are not yet available in Google Sheets or Google Meet, and currently support English prompts and instructions only. Admins can monitor and audit skill activity in the Admin console and disable the feature at the domain, organizational unit, or group level if needed.

Sources :Workspace Updates Blog

Recherche

AWS Expands STS to Support Cross-Partition Temporary Credentials, Replacing Long-Term Keys

Sujet · AWS跨分区临时凭证Prise rapide
2026-11-10 15:08 GMT+8

Amazon Web Services (AWS) has announced an expansion of its Security Token Service (STS), enabling workloads to use their existing identities to obtain temporary credentials across different AWS partitions, such as the Commercial partition and the GovCloud (US) partition. This update aims to address the need for enterprises to store long-term IAM access keys for cross-partition operations, thereby reducing the risk of key leakage and management overhead.

AWS partitions are independent environments composed of one or more cloud regions, each with its own Identity and Access Management (IAM) system. Previously, credentials issued in one partition could not be used directly in another. The new mechanism utilizes existing GetWebIdentityToken and AssumeRoleWithWebIdentity APIs: workloads first obtain a short-lived JSON Web Token (JWT) from the source partition's STS, then send it to the target partition's STS to exchange for temporary credentials for a specified IAM role. The target partition's STS validates the token and issues credentials based on the IAM role's trust policy, with access still restricted by IAM permission policies.

Cross-partition access is not enabled by default; administrators must enable external identity federation in the source account and establish trust relationships in the target account. Enterprises can set access conditions based on source accounts, IAM identities, and regions, with AWS CloudTrail logging token issuance and credential exchange activities in both partitions. The feature is currently available in all AWS Commercial Regions, GovCloud (US), China Regions, and the AWS European Sovereign Cloud region in Germany at no additional cost.

Sources :iThome

Recherche

Japan's LINE Yahoo Launches Standalone iOS App for AI Agent 'Agent i' with Price-Drop Alerts

Sujet · Agent iPrise rapide
2026-11-10 15:23 GMT+8

Japanese internet giant LINE Yahoo Corporation launched a standalone iOS application for its AI agent "Agent i" on November 10, 2026. Previously accessible only within the LINE or Yahoo! JAPAN main apps, the service can now be launched with one tap from the home screen. An Android version is expected to roll out subsequently.

Agent i was introduced in April 2026 by integrating Yahoo! JAPAN's "AI Assistant" and LINE's "LINE AI." The new standalone app features a "Task Function" that allows users to register keywords for continuous monitoring of latest news, product price drops, or event schedules, sending app notifications when tasks are completed. It also supports quick access to new chats, image input, and voice input via home screen widgets.

The number of "Domain Agents" has expanded to a cumulative 40 (including beta versions), featuring an "Oshikatsu" (fan support) agent launched on September 24. This agent organizes the latest information, events, tickets, and merchandise for users' favorite artists or characters.

Users can log in with a LINE account or Yahoo! JAPAN ID; some functions are available without login, though free-form text input requires authentication. According to reports, Agent i utilizes the Gemini Enterprise Agent Platform (formerly Vertex AI) or OpenAI APIs under the hood. The app is rated 13+ on the App Store, with guidelines advising against use by children under 13.

Sources :テクノエッジ TechnoEdge

Recherche

Incoming Texas Governor Warns AI Firms: The Data Center 'Free-for-All' Is Over

Sujet · 德州数据中心监管Matériel
2026-11-10 09:00 GMT+8

Texas Governor-elect Gina Hinojosa has warned artificial intelligence companies that the era of unchecked data center expansion in the state is over. In an interview with CNN, Hinojosa stated that data centers have moved into rural communities without rules or laws to protect Texans, asserting that "Texans are being taken advantage of."

She pledged to introduce legislation to reverse this trend, calling for a pause on new data center developments until protections are in place. "We need to stop what we are doing with data centers until we have those protections," Hinojosa said. She emphasized that communities must have a seat at the table to decide if and how these facilities enter their areas. "It is not that I hate data centers. I hate the way they are taking advantage of Texas, and that stops when I'm Governor."

Texas has long been one of the most data center-friendly locations in the US, but recent polls show 75% of respondents oppose building such facilities near their homes. With Democratic candidates winning key races in the mid-term elections, including Hinojosa's victory for governor, there is growing political momentum against rapid AI infrastructure growth.

This shift could impact plans for major AI firms like Anthropic, whose recent filings noted that delays or failures in their complex data center supply chains could significantly affect the company.

Sources :TechRadar

Recherche

Investment Blogger Klement: The Funding Gap Between US and European Startups Concentrates at the Super Unicorn Stage, with Similar Growth-Stage Expansion Rates

Sujet · 欧美初创融资差距Matériel
2026-11-10 15:09 GMT+8

The most intuitive manifestation of the funding disadvantage facing European startups is the stark contrast in financing scale: Mistral’s $3.5 billion Series D round stands as the largest venture capital round in European history, whereas OpenAI secured $122 billion in March 2026, Anthropic raised $65 billion in May 2026, and xAI obtained $20 billion in the first quarter of 2026. Placing Mistral’s $3.5 billion round within a list of the largest US funding rounds would rank it only 19th. Investment blogger Joachim Klement uses these figures to pose a question, but subsequently cites a new OECD study to sharpen the comparison.

The OECD study covers startups founded in the European Union and the United States between 2000 and 2025, categorizing successfully scaled enterprises into two types: 'rising superstars' (companies eventually reaching valuations over $11 billion) and 'growth-oriented companies' (those raising lead rounds of $50 million). The core finding is that growth-oriented companies experience roughly identical financing growth rates in both Europe and the US—indicating that Europe does not lack capital supply for mid-sized growth firms. The real divergence occurs at the super unicorn stage: five to six years after launch, US firms begin attracting far greater capital than their European peers.

The OECD also provided another set of data from a temporal perspective: from the first patent application to securing sufficient funding for scaling, approximately 40% of EU startups took more than seven years, averaging 1.5 years longer than their US counterparts. Klement presents these two findings side by side to illustrate that European startups face not only a gap in top-tier capital attraction but also a delayed start in the entire scaling process. The original OECD report could not be independently verified; the above data are all paraphrases by Klement of that research.

For investors, the practical implication of this finding lies in improving diagnostic precision. If one assumes Europe’s problem is a generalized capital shortage, then any early-stage or growth fund faces the same dilemma. However, if the issue concentrates at the super unicorn stage, the variables requiring attention differ: it is not whether Europe has venture capital, but whether it has sufficiently sized follow-on round institutions and deep exit markets to support companies transitioning from good to exceptional. Klement teased at the end of the article that he would discuss another study on Thursday, identifying another key factor preventing European startups from scaling, though this was not elaborated upon in this piece.

Sources :klementoninvesting.substack.com

Recherche
Suivant →
Connexion / Inscription
LectureRechercheRadarCadre d'investissement

Dyson Chief Engineer Jake Dyson previously recalled in an interview that Dyson has historically been very strong in mechanical engineering, motors, airflow, and performance, and that now, with the addition of cameras and visual systems, the machine can further detect what it sees, understand the environment, and act accordingly. This statement highlights the strategic intent behind the product: when powerful mechanical engineering is supported by perception and judgment, the device gains the intelligence to solve problems independently.

The same "perception-judgment" mechanism, the author believes, can handle home floors as well as identify hair quality, sound, oral environments, and even broader fields. CameraJet toothbrushes with cameras and more new products with sensing capabilities are emerging.

It should be noted that figures such as the maximum suction (30,000 Pa), number of recognized types (nearly 200), and dust bin emptying cycle (100 days hands-free) are manufacturer specifications or subjective observations during the editor's short-term use, not independent benchmark test results. Ifanr's practical testing duration has not yet reached the upper limit of the emptying cycle, nor has it conducted same-caliber comparisons with competing products. These limitations do not affect the discussion of the design logic itself, but readers should view them as manufacturer-reported capabilities attributed to a single editor's experience when evaluating actual cleaning effectiveness.

The next-generation Xeon processor, Diamond Rapids, utilizes Intel's 18A-P process technology, offering up to 256 cores, 1.28 GB of last-level cache, 16 memory channels, and 128 lanes supporting PCIe Gen 6 and CXL 3.0, along with Foveros Direct 3D and UCIe-S interconnects. Intel claims that compared to 18A, 18A-P improves performance by approximately 9% at the same power consumption, or reduces power consumption by approximately 18% at the same performance level. However, this is a comparison at the process node level and cannot be directly equated to improvements in an entire server.

Whether these specifications can translate into stable supply, mature software, and prices that customers are willing to adopt remains the next test. For enterprises, the actual basis for evaluating hardware is how much time, electricity, and cost are required to complete each task.

The root cause was that the Agent at that time possessed only static code analysis capabilities; it could not prepare environments, set breakpoints, control thread execution order, or inspect memory data. Its explanations did not match the program's actual runtime behavior, making independent validation impossible. To solve this problem, the team added two types of tools: debugging and memory analysis.

Regarding debugging, the core components were written in Java. The team initially tried JDB, the debugger included in the JDK, but it was designed for human interaction, requiring the maintenance of terminal processes, sending commands via interprocess communication, and reading outputs, which was cumbersome for an Agent. They reimplemented the JDB CLI based on the Java debugging interface, allowing the tool to keep a debugging session and state alive in the background. Each foreground command would exit after execution, and the next invocation would continue operating on the same session. Results were returned in structured form for the Agent to parse. A key capability was thread-specific breakpoints: only the target thread could trigger a breakpoint, and when hit, only that thread was suspended while others continued running. Combined with resume operations, this allowed controlling relevant threads to execute alternately in a specific order.

Regarding memory analysis, by examining heap memory snapshots (Heap Dumps) of Java programs, one can check field values, reference relationships, and memory usage of relevant objects. Heap Dump files are usually very large, so the team placed them on cloud servers with large memory for centralized loading and processing, providing specialized query interfaces for the Agent. The Agent queries memory data on demand and continues analysis based on the results, without needing to process the entire large file locally.

With tools in place, the Agent also needed a runtime environment. The team provided two options: PolarDB-X Zero can create temporary instances within seconds, suitable for quickly reproducing simple issues like SQL errors or abnormal functionality; if a specific historical version, specific topology, or complex concurrency scenario is required, a full reproduction environment is built in K8s, matching the version and topology of the affected instance. After completing code modifications, the Agent can also deploy the fixed version to this environment and use the same method to check whether the original problem has been resolved.

Returning to that concurrency bug, this time the Agent could attempt reproduction itself, continuing analysis and debugging based on actual execution results. After several iterations, it identified the key triggering conditions and reproduced the same MDL lock-leak state seen in the original issue. The key was that the Agent used debugging tools to make two threads execute alternately in a specific order, deliberately recreating an intermittent issue that had previously been difficult to encounter. Debugging results could also be cross-validated against memory states to confirm that the reproduced issue was indeed the same one. At this point, both root cause analysis and a repeatable sequence of operations were available, and the Agent's analysis was supported by evidence from actual execution.

In the delivery phase, the team made reliably reproducible test cases a key threshold for entering code review. The fact that the Agent added tests and passed them is not sufficient to indicate that the original problem has been solved; it must first be confirmed that the test case actually triggered that problem. The same test case must reliably reproduce the original problem on the pre-fix version and verify that the problem no longer appears on the post-fix version. Only after this before-and-after validation is complete and full regression tests pass can the code enter review.

After this workflow ran for six months, the team identified and recorded over 200 defects through processing tens of thousands of alerts, of which over 70% could be reliably reproduced and fixed by the Agent before entering code review. All alerts first underwent initial analysis and handling by the Agent, achieving 100% first-round coverage.

Sheng Taotao summarized three transferable engineering experiences: First, make development systems accessible to the Agent, starting with systems frequently used by the team, equipped with CLI tools or exposed API interfaces; Second, make the development process verifiable, providing debugging tools and runtime environments so the Agent can reproduce problems and validate its analysis itself, where feedback from actual execution is key to discovering errors, adjusting analysis, and ultimately completing the fix; Third, enable the task workflow to run continuously, moving the workflow to the cloud so the Agent does not depend on developers' local computers.

It should be noted that all the above data are self-reported by the PolarDB-X team and have not been independently verified. This method applies to defect repair scenarios where the problem definition is clear and code changes are small; for new feature development requiring developers to lead solution design, the assisted coding mode is still adopted. Building custom CLI tools and tailored debugging interfaces requires certain infrastructure investment, and the cost of migrating to other teams depends on the accessibility of their existing systems.