Create

Sign in to ReadmeX

or

How we handle your data: Privacy Policy

ReadmeX

Your AI reading companion

What would you like to catch up on?

Pick a briefing, or ask your own question.

Your community and people briefings stay personal to you.

Huawei

AI overview

Sign in and the AI will write an overview from our coverage.

Headlines · 239

  1. Chinese tech firms reportedly bought over 1 million Nvidia GPUs for AI

    The Financial Times, citing industry sources, reported that Chinese technology companies bought more than 1 million Nvidia gaming and workstation chips this year, mainly RTX 5090 and RTX Pro models, to address shortages in AI computing capacity. The report said purchases were concentrated on export-compliant RTX Pro 6000D, RTX Pro 5000 and RTX 5090 cards, which are better suited to simpler workloads such as prefill than advanced AI tasks. It also cited a preference for Nvidia's mature CUDA platform, whose support across AI frameworks and libraries and use in existing code made it attractive despite Huawei's CANN alternative. Another report said Chinese customers had ordered more than 6 million RTX Pro 5500 cards.

    Techmeme · 🔥 0
  2. IDC: Nvidia's China AI chip share falls to 34% as Huawei rises to 28%

    IDC data reported by Reuters show Nvidia's share of China's data-centre AI chip market fell to 34% in the first half of 2026, with 678,000 AI accelerators shipped, down 44% year on year. Huawei shipped 572,000 accelerators, up from more than 300,000 a year earlier, to reach 28%, while nine of the 11 Chinese chipmakers tracked by IDC grew shipments; Alibaba's T-Head held 7%, Cambricon 5%, Baidu's Kunlunxin 4% and Hygon 3%. IDC did not specify which products made up Nvidia's 34% share or whether B200 and B300 chips were included; cnBeta reported total China AI accelerator shipments of 2.008 million, with GPUs down 26% to 991,000 and other accelerators up nearly 80% to 1.016 million.

    Techmeme · 🔥 0
  3. Wang Yuquan: AI is no bubble, but 2029 is the risk window

    Technology investor Wang Yuquan of Haiyin Capital argues in a conversation that the current AI boom is not a bubble but a decade-plus technology shift, with the real risk window around 2029, when application paradigms in AI software, AI coding and autonomous driving become clear while revenue lags behind infrastructure spending. He expects 24/7 trading, quantitative strategies and stablecoin growth to amplify market swings, hurting memory makers such as SK Hynix and South Korea hardest, while leaders like Nvidia are undervalued. He also calls embodied AI and large language models overhyped, urges Chinese labs to close their models for safety and revenue reasons, and predicts a 70-30 split favoring China outside the US.

    虎嗅 AI · 🔥 0
  4. HarmonyOS 7 Xiaoyi AI shortcut bar expected to roll out from late October

    Huawei customer service told IT之家 that the Xiaoyi AI shortcut bar, a new HarmonyOS 7 feature, is expected to begin rolling out in late October, pushed out in batches by device model. The feature was first teased at the September 7 launch event for HarmonyOS 7 alongside the HUAWEI Mate XT 2 and other products, where official slides showed Xiaoyi taking up permanent residence in the dock at the bottom of the phone home screen once enabled. HarmonyOS 7 runs on the Harmony Intelligence architecture and supports the system-level agent Xiaoyi, combining a four-layer memory architecture, app Skills, cloud and on-device large models, and multi-device collaboration.

    IT之家 AI · 🔥 0
  5. China's AI boom shifts from token prices to power bills

    NextFin News, published via TMTPost, reports that Ulanqab in northern China has become one of the country's densest data-center clusters, with Alibaba, Huawei, Kuaishou and specialist operators taking space because of relatively cheap electricity, available land and a climate that eases cooling loads; one abandoned school is being converted into a computing facility. The building boom coincides with falling model prices: DeepSeek made a 75% cut to its V4-Pro API rates permanent in May, and a Token price index compiled by Silicon Data and cited by the Financial Times showed composite prices more than 52% below their May peak by October, though not every model fell equally. Compute sellers are still spending more — Alibaba's capital expenditure for the quarter ended June 30 reached 67.678 billion yuan, up 75% year on year, while AI cloud and computing services revenue rose 45% to 48.437 billion yuan — and the article's hypothetical example of a 10-megawatt IT load at PUE 1.2 implies a roughly 42 million yuan annual electricity-cost gap for every 0.40 yuan per kWh difference, over 200 million yuan cumulatively across five years.

    钛媒体 · 🔥 0
  6. Huawei details Kirin 9050 Pro, its first logic-folded chip, with on-device 30B MoE

    At a 2026 Kirin media briefing in Shanghai on Oct 13, Huawei semiconductor executive Xu Songyu published the full architecture diagram of the Kirin 9050 Pro, the first logic-folded "Tao" chip, which Huawei says has been mass-produced with support from China's semiconductor industry. The chip combines a new Lingxi CPU (9 cores, 16 threads), a fourth-generation Maleoon GPU with second-generation hardware ray tracing (40% higher performance, 45% lower power, 78% better ray tracing), a new Da Vinci NPU, a Kirin AI ISP, a Balong AI Modem, an FFTS function-flow task scheduler, a security processor and the Lingxi micro-core. Huawei says the NPU raises AI performance by 140% while cutting power by 66% and enables the industry's first on-device 30B MoE model (2B active parameters, under 2.5GB of memory), alongside 7B dense and 1B VLM models; Xu also said the next-generation Kirin 2027 chip is ready and that all test results so far meet expectations.

    IT之家 AI · 🔥 0
  7. Huawei-ECNU time-series model tops TIME-leaderboard and GIFT-Eval

    According to Huawei Cloud, a time-series pretrained model called Qiyao, developed jointly by Huawei teams (Huawei Cloud industry large-model team, the 2012 Applied Scenario Innovation Lab and others) and East China Normal University, has topped two international time-series benchmarks: TIME-leaderboard and GIFT-Eval (Pretrained category), as reported by ITHome. Huawei Cloud said the team deeply optimized training and inference on the Ascend ecosystem, yielding a model it describes as highly accurate, robust, general-purpose and easy to engineer. The company says the model can be used directly for zero-shot time-series forecasting in areas such as grid load assessment, industrial equipment anomaly detection and financial trend analysis, and supports low-cost fine-tuning for vertical industries.

    IT之家 AI · 🔥 0
  8. Five Chinese flagship launches in 11 days put on-device AI at the center

    A roundup republished by Cyzone (originally from Lansha Finance) describes how vivo's X500 series, OPPO's Find X10 series, Xiaomi's 18 Pro series, Honor's Magic9 series and Huawei's Mate 90 series all launched within 11 days, from Sept 21 to Oct 1, earlier than in past years. The piece argues the competition has shifted from pixel and chip specs toward imaging, on-device AI and long-term system experience: Xiaomi 18 Pro adds a dedicated AI button and AI auto-editing, vivo and OPPO embed AI in OriginOS 7 and ColorOS 17, Huawei folds AI into HarmonyOS, and Honor turns YOYO into a system-level agent while announcing a MobileAI plan with China Mobile that bundles AI token allowances with telecom plans. It cites Counterpoint Research projecting generative-AI phones at 45% of 2026 global shipments, while noting AI is not yet a strong enough reason for users to upgrade, and notes 7,000mAh-plus batteries, 200MP cameras and 8K/Log video as shared trends. Sales and ranking figures (such as Honor's claim of 208% first-hour growth and leading Android flagships during the holiday) come from the vendors, while pricing and DRAM data are attributed to IDC and TrendForce.

    创业邦 科技 · 🔥 0
  9. GAC Aion Ray7 to open pre-sales Oct 17 with Huawei e-drive, CATL battery

    GAC Aion's Ray7 will open pre-sales on October 17, according to a report relayed by Yiche, with the car built on the AEP pure-electric platform, using a rear single motor, Huawei's Drive ONE multi-in-one electric drive and CATL batteries. Cabin hardware includes a Qualcomm 8295 processor, 24GB LPDDR5 memory and 128GB UFS5.0 storage; the company claims it is the first to offer Huawei EMB fully electric brake-by-wire plus Huawei iTRACK torque control, in a body measuring 4960×1900×1455mm with a 2920mm wheelbase. For driver assistance, the report says it uses an algorithm derived from WeRide's L4 model along with 27 automotive-grade sensors, including a 192-line lidar and three 4D millimeter-wave radars, supporting features such as finding a parking space without navigation.

    IT之家 AI · 🔥 0
  10. China's AI industry shifts from buying compute to running it

    The Financial Times reported on October 8 that Ulanqab in Inner Mongolia has 89 signed data-center projects, with Alibaba, Huawei, Kuaishou and third-party operators building there, drawn by cheap power, land and a climate suited to server cooling; the city aims to become a "capital of tokens." Prices are falling at the same time: Reuters reported in May that DeepSeek made a 75% cut to its V4-Pro API pricing permanent, taking tiers from 0.1–24 yuan per million tokens down to 0.025–6 yuan, while the FT in October cited Silicon Data's token price index showing composite token prices more than 52% below their May peak, though not all models fell equally. The article pairs Alibaba's 67.68 billion yuan quarterly capex (up 75% year on year) with 48.44 billion yuan in AI cloud and compute revenue (up 45%), arguing that China's AI competition is moving from acquiring compute to operating it, with depreciation and payback now the hard questions.

    虎嗅 AI · 🔥 0
  11. Huawei opens public beta of three DevEco tools for HarmonyOS PC

    According to Zhidx, Huawei on Oct 8 opened public beta of three HarmonyOS PC development tools: the visual IDE DevEco Studio (moving from invite-only testing), the AI development tool DevEco Code, and the AI capability suite DevEco CLI. The report describes a division of labor — DevEco Studio as the integrated development and multi-device debugging base, DevEco Code offering three layered modes (Goal, Plan, Build) for end-to-end agentic development, and DevEco CLI packaging the HarmonyOS toolset, official knowledge base and 70-plus Skills into command-line interfaces that AI agents can call. The report says HarmonyOS 6 and 7 hardware devices have passed 90 million units, driving demand for apps and agentic services.

    智东西 · 🔥 0
  12. MemoraX pitches memory-model-driven recursive self-improvement for agents

    Machine Heart reports that MemoraX AI has introduced a "Memory Model–Driven RSI" approach in which a dedicated memory model distills and updates procedural memory from an agent's execution traces and online feedback, while the agent system applies that experience to new tasks, forming a distill–apply–verify–update loop. The company says the method raised pass rates by more than 20% and cut the main coding agent's token use by more than 20% across 13 industrial long-horizon coding tasks (78 runs, over 335 hours total), and that 36 iterations on the financial benchmark FinRank lifted task accuracy from 33.33% to 73.33% with 96.49% evidence recall and roughly 98% lower cost than a Codex baseline. The piece also cites OpenAI's stated goal of automated AI researchers and Anthropic's claim that Claude already leads about 26% of its internal AI R&D work; all performance figures come from the vendor and have not been independently verified.

    机器之心 · 🔥 0
  13. CNCC 2026 set for Oct 21–24 in Chengdu with 12 academicians, 789 speakers

    Organizers announced at a Oct 10 Beijing press conference that the 23rd China National Computer Congress (CNCC 2026) will run Oct 21–24 in Chengdu under the theme “Data Drives, Intelligent Computing Shapes the Future.” The program includes 17 invited talks, 3 plenary forums, 167 thematic forums and over 40 side events, with 789 speakers — among them 12 academicians and 92 experts from companies including Tencent, Alibaba, Baidu, Xiaomi and JD.com. Tracks newly include “AI + industry,” split out agents and data science/data spaces, and merge vision, multimodal and world models.

    智东西 · 🔥 0
  14. Huawei says Ascend tops Nvidia in China share as supply stays tight

    According to a media briefing summary published by Jiemian, Huawei rotating chairman Xu Zhijun said Ascend is still far from meeting domestic demand and that Huawei has no plan to expand broadly overseas, urging the supply chain to "go faster, produce more"; he said Huawei's own data suggests Ascend's China market share has surpassed Nvidia's, though the note gave no figure or methodology. Nvidia said in its quarterly report that H200 shipments to some Chinese customers permitted since February accounted for less than 1% of data center revenue in the quarter, and it excluded China data center compute revenue from next-quarter guidance. The article also reports intent prices for Ascend 950DT accelerator cards up 20%-50% in two months to over 250,000 yuan and 950PR up from about 60,000 yuan early this year to over 80,000 yuan (intent quotes, not final prices), and that DeepSeek published Ascend-targeted software components including TileKernels and DeepGEMM-Ascend on September 30.

    虎嗅 AI · 🔥 0
  15. Tencent-backed AI storage vendor XSKY refiles for Hong Kong IPO

    Chinese enterprise AI storage provider XSKY (Beijing Xingchen Tianhe) has updated its prospectus for a second attempt at a Hong Kong main-board listing, after eight funding rounds whose investors include Tencent, Boyu Capital, Qiming Venture Partners and Redpoint China; its post-F+ round valuation was about RMB 4.58 billion. The filing shows the company only turned profitable in 2025, with RMB 7.137 million of profit and just RMB 307,000 in the first half of 2026, while appliance gross margin fell from 25.5% to 18.9%, R&D spending dropped from 65.9% to 22.3% of revenue, and operating cash flow stayed negative. XSKY ranked second in China's distributed AI storage market with a 10.6% share by 2025 installed base, but founders Xu Xin and Wang Haomai face share-redemption obligations if the listing fails.

    钛媒体 · 🔥 0
  16. AI phone entry war: Huawei, Alibaba, ByteDance and Tencent chart four paths

    An analysis piece published by TMTPost lays out how China's biggest tech players are approaching AI phones: Huawei's Mate 90 series, launched Oct 1, is described as running a 30B MoE omni-modal model on-device via the Kirin 9050 Pro; Alibaba pairs its Qwen Intelligence stack with Honor for device-cloud co-development; ByteDance and Nubia launched the second-gen Doubao phone, NaviX Ultra, billed as the "world's first AI agent phone"; and Tencent avoids the system-level agent race, instead plugging into Huawei, Xiaomi, Honor, OPPO and vivo through WeChat's A2A protocol while working with Qualcomm to deploy Hunyuan 7B/3B on-device. The article cites an IDC forecast of 147 million next-gen AI phones shipping in China in 2026, up 31.6% year on year and 53% of the market.

    钛媒体 · 🔥 0
  17. Qualcomm to pay Huawei under patent cross-license, says analysis

    An analysis published by TMTPost (authored by Shi Tianhao) says Huawei and Qualcomm reached a multi-year, broad patent licensing agreement covering cross-licenses in 5G, computing, AI and networking, with Qualcomm also acquiring some of Huawei's US patents in computing, AI and networking. Citing CCTV Finance, it says Qualcomm will pay to use more than 680 Huawei patents, and that Huawei's cumulative licensing revenue across all agreements is expected to exceed $6.9 billion (the value of this deal was not disclosed); Qualcomm has denied being a "net payer" and denied the transaction involves "logic folding." The piece sets the deal in a 25-year timeline in which Huawei, once the payer, has become a payee, contrasts the Kirin 9050 Pro's "logic folding" and Taifa with Xiaomi's XRING O1, and notes claims that Ascend's China market share has passed Nvidia, that DeepSeek is co-developing open-source software tooling for Ascend, and that AMD is buying World Labs for about $8.2 billion.

    钛媒体 · 🔥 0
  18. Huawei weighs taking HarmonyOS global as device base tops 90M

    Yu Chengdong, Huawei's consumer business chairman, told international media that Huawei is considering gradually bringing its HarmonyOS operating system to global markets, though he called it an initial plan with no timetable. He said more than 90 million devices now run HarmonyOS 6 and HarmonyOS 7, with 100 million expected by the end of 2026, and warned that sharply higher memory prices have added more than $200 to the cost of each phone, forcing Huawei to raise prices. Huawei also signed a multi-year patent cross-licence with Qualcomm covering 5G, computing, AI and networking, subject to regulatory approval.

    钛媒体 · 🔥 0
  19. AI's memory squeeze: DRAM prices triple as PC and phone costs climb

    Merchants in Shenzhen's Huaqiangbei market say memory module prices have risen more than 300% in a year, with 16GB DDR5 kits going from just over 200 yuan at their cheapest to 1,400 yuan or more, and 16GB DDR4 from about 100 yuan to over 600 yuan. The piece argues that AI servers need 8–10 times as much memory as standard servers, that roughly 70% of advanced memory capacity is being consumed by AI, and that Samsung, SK Hynix and Micron keep shifting output to higher-margin HBM, structurally shrinking general-purpose DRAM supply. It also notes DRAM die costs rose only about 30% while end prices climbed 200–300%, attributing part of the gap to channel stockpiling and AI-themed speculation; phone and PC makers have raised prices, and Gartner expects average PC prices to rise 17% by the end of 2026, with sub-$500 entry-level PCs possibly disappearing by 2028.

    钛媒体 · 🔥 0
  20. openJiuwen releases and open-sources enterprise AgentOS

    openJiuwen, an open-source AI agent platform built by teams including Huawei's 2012 Labs, Huawei Cloud and others together with universities and outside developers, has released and open-sourced an enterprise-grade AgentOS. It bundles multi-agent swarm collaboration and workflow orchestration, self-evolution, compute affinity, and enterprise features such as multi-tenant isolation, security sandboxes, permission management and failure recovery into one base layer with a plugin-based architecture. The article says skills, connectors, plugins, experts and expert teams can be published and shared through Agentic Hub, and that at Huawei Connect 2026 Huawei also announced an all-in-one open Agent acceleration platform built on Jiuwen AgentOS with the WorkSwarm office/programming workbench and a SystemAgent, claimed to enable hour-level end-to-end private deployment. Code is hosted on AtomGit and GitHub.

    量子位(原生 RSS) · 🔥 0
  21. Smartphone launch wave: higher prices, on-device AI agents, foldables going mainstream

    An industry analysis by Qingcheng Caijing describes a dense run of flagship launches around September 2026 from Huawei, Xiaomi, Apple, vivo, OPPO, Honor and Nubia, with Chinese flagship starting prices rising and foldables reaching the 10,000-yuan bracket. It attributes the price hikes to AI compute consumption of HBM crowding out LPDDR supply for phones, and says vendors are moving from decorative AI features toward system-level agents such as Honor's YOYO Harness, vivo's device-cloud agent matrix, and Nubia's NaviX Ultra with the Doubao phone assistant. The author argues foldables are the only hardware innovation currently supporting premium pricing, with value shifting to large-screen AI multitasking and ecosystem coordination rather than the fold itself.

    虎嗅 AI · 🔥 0
  22. From 8-GPU servers to supernodes: AI data centers for the inference era

    Speaking at the 2026 ITValue Summit on September 17, xFusion computing vice president Wang Liangdong said AI has entered an inference-dominated phase in which global token consumption keeps surging and buyers judge infrastructure by cost per token, throughput and cluster stability rather than peak per-card performance. The article cites a range of figures: the 2026 China Development Forum said China's daily token calls topped 140 trillion in March, OpenRouter data showed China accounted for 36% of global token calls in the March 16-22 week, JPMorgan expects Chinese inference token consumption to grow roughly 370x in five years, and IDC projects annual global consumption of 150,000 Peta Tokens by 2030. It argues the 8-GPU server paradigm hits communication bottlenecks under trillion-parameter, high-concurrency inference, pointing to Huawei's Ascend 960 supernode (4,096 cards per node, 8E FP8), Lenovo's Wanquan heterogenous computing platform V5.0 and xFusion's FusionPoD for AI rack-scale designs — while warning that supernodes only suit very large workloads, interconnect standards remain unsettled, and utilization depends on sustained paying demand.

    钛媒体 · 🔥 0
  23. AI is becoming the front door to healthcare, from Doubao to Ant AQ to ChatGPT Health

    The piece traces AI's shift into the first stop for health questions: by March 2026 China's AI-native apps reached 446 million monthly users, with Doubao at 345 million, while Ant launched standalone AI health app AQ in June 2025 and by 2026 reported more than 100 million users and over 10 million health consultations a day. OpenAI said more than 230 million people a week asked ChatGPT health questions when ChatGPT Health launched in January 2026, rising past 300 million by July. The article stresses that AI handles pre-visit questioning and record organisation while diagnosis and treatment advice stay with doctors — a line both JD Health and OpenAI explicitly keep.

    36氪 人工智能 · 🔥 0
  24. Nikkei: China's major AI firms launched 16 new models in September

    Nikkei data reported by TechNews shows 10 major Chinese AI companies — including DeepSeek, Alibaba, Tencent and Xiaomi — launched 16 new models in September, mostly lightweight and low-cost, such as DeepSeek V4.1-Flash, Zhipu's GLM-5.3-FlashX and MiniMax M3.1-Flash-Preview. BenchLM counted 20 large language models from eight major Chinese firms in July–September, up from 2 in the first quarter and 17 in the second, while the anonymous model Space Bunny Alpha — believed to be a stealth MiniMax release — surged in usage on OpenRouter. The report contrasts this with Anthropic CEO Dario Amodei's calls to slow AI R&D and crack down on illegal distillation, noting Chinese firms increasingly train and run models on domestic chips.

    科技新报 · 🔥 0
  25. Clip-on earbuds up 82% as AI agents fail to sell hardware

    A Baobian piece republished by 36Kr says IDC data shows global shipments of clip-on open-ear earbuds jumped 82.2% year-on-year in Q2 2026, while ear-hook models fell 5.1%; in China, headphone retail sales dropped 15.9% to 87.97 million units in H1 2026 even as open-ear devices grew 34.1%. The article argues the category's rise, with no deep agent or large-model features, shows that comfortable, low-effort basics still win buyers, and that AI features are overrated as a trigger for hardware upgrades.

    36氪 人工智能 · 🔥 0
  26. Samsung joins Nvidia and KKR in $1B Helix Digital Infrastructure deal

    Samsung and five affiliated businesses are committing a combined $1 billion to Helix Digital Infrastructure, a data center venture founded by KKR in June 2026, with Samsung Electronics contributing $500 million. Helix's founding investors also include Nvidia, the Kuwait Investment Authority and US power company Vistra, and the company plans to secure power capacity and network connections for its sites. Samsung says the arrangement moves it from supplying hardware toward directly participating in AI infrastructure build-out.

    TechRadar · 🔥 0
  27. DeepSeek open-sources Ascend ports of DeepGEMM and DeepEP kernels

    DeepSeek has published two new repositories, DeepGEMM-Ascend for matrix multiplication and DeepEP-Ascend for the all-to-all communication used by mixture-of-experts models, alongside Ascend-targeted updates to TileLang, TileKernels, DeepSelect and FlashMLA. The DeepGEMM changelog dates the Ascend release to 2026.09.30; DeepGEMM itself is a tensor-core kernel library covering FP8, FP4 and BF16 GEMMs, fused MoE and MQA scoring, with all kernels compiled at runtime through DeepJIT. SemiAnalysis reported that at the time of writing only TileLang was usable, because DeepSeek requires CANN 9.2.0 and a C++20 compiler that the beta toolkit did not provide, while TechRadar said DeepGEMM-Ascend needs Ascend 950 series silicon and CANN 9.20. The README's claim that performance matches or exceeds expert-tuned libraries across matrix shapes is the maintainers' own and has not been independently measured.

    GitHub Trending(每日) · 🔥 0
  28. Qualcomm reportedly to license Huawei patents tied to LogicFolding

    Qualcomm has reportedly agreed to a multiyear cross-license with Huawei that may include patents supporting Huawei’s LogicFolding chip-stacking architecture, Bloomberg reported. The companies’ announcement did not identify the patents or mention LogicFolding; the deal also includes technology areas such as 5G, AI services, optics and networking, and remains subject to regulatory approval.

    Tom’s Hardware · 🔥 0
  29. MoWorld world model reaches Huawei Mate 90 phones

    Moxin Technology is reported to have deployed its MoWorld-2.0-Flash world-model variant in the HarmonyOS-exclusive Moxiao Pet app for Huawei’s Mate 90 series. The source says cloud infrastructure generates 3D Gaussian assets while the approximately 600-million-parameter model runs on the phone for rendering, but the performance and cost claims are promotional and unverified.

    新智元 · 🔥 0

← Previous

Experience and discussion from the community

Share my Huawei experienceAsk about Huawei

vllm-ascend: A community-maintained hardware plugin for running vLLM on Huawei Ascend NPUs.

What problem it solves vLLM users need a way to run the framework on Huawei Ascend NPU hardware through a supported integration path. The plugin decouples Ascend NPU integration from vLLM and enables supported model types to run on the platform. Key features Hardware-pluggable vLLM integration — The plugin provides an interface for integrating Ascend NPUs with vLLM and follows the community's hard…

github.comhttps://github.com/vllm-project/vllm-ascend