Alibaba unveils Zhenwu V900 AI chip, targets 20GW data centers and 10-trillion-parameter model by 2032
At the 2026 Apsara Conference in Hangzhou, Alibaba Group CEO Wu Yongming announced a major AI infrastructure push, unveiling the Zhenwu V900 AI chip with triple the performance of its predecessor, plans to train a 5-10 trillion parameter model, and a target for Alibaba Cloud to operate over 20GW of global data center capacity by 2032. The chip is set for mass production in Q1 2027.
IllustrationEditorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page reads the event directly, while its address stays stable when the title changes.
- Summary covers the current reports
Cross-source coverage
Common ground
- Alibaba's AI infrastructure buildout is a strategic move in a geopolitical context, not just a business decision.
- The US export controls on chips have pushed China to develop domestic alternatives like the Zhenwu V900.
- The Global South is watching closely for alternatives to Western tech dominance.
- There is significant execution risk in scaling from under 10% to 40-50% self-developed chips while building 20GW of data centers.
- The capex-to-revenue math is concerning, with cloud revenue growth at 10-15% versus massive spending.
Points of contention
- Whether Alibaba's strategy is a sound business investment or a politically driven gamble with questionable returns.
- Whether the Zhenwu V900 chip can truly compete with Nvidia's latest or is only impressive within China's domestic context.
- Whether Alibaba's infrastructure serves the Global South as a liberating alternative or just a new form of dependency.
- Whether China's AI governance provides accountability or enables surveillance and control without oversight.
- Whether the 650 enterprise customers represent genuine market demand or government-mandated adoption.
Blind spots
- The debate overlooks the environmental cost of 20GW of data centers in a country with severe water shortages.
- No one has addressed how Alibaba's infrastructure will affect ordinary people's lives, not just corporations or states.
- The question of who pays if the investment fails—shareholders, taxpayers, or consumers—is ignored.
- The interoperability needs of developing nations, who want choice between US and Chinese ecosystems, are not fully considered.
- The potential for Alibaba's technology to become obsolete before it's monetized is not discussed.
WorldAttention’s read
Alibaba's AI infrastructure plan is a bold, strategic bet that blends technology, geopolitics, and governance. While it offers a potential alternative to Western dominance for the Global South, the financial math is shaky—massive capex with uncertain revenue growth. The debate reveals deep divides: some see it as a necessary step for sovereignty, others as a risky, politically driven project with accountability gaps. The real blind spots are the environmental costs, the impact on ordinary people, and whether developing nations will gain genuine choice or just a new master. Ultimately, the success hinges on Alibaba's ability to monetize its compute before it becomes obsolete, a question that remains unanswered.
Reporting timeline
Alibaba CEO predicts machine thinking will dominate; DeepSeek hires 90s CFO ahead of IPO
At the 2026 Alibaba Cloud Computing Conference, Alibaba Group CEO Wu Yongming outlined a vision where machines become the primary source of thinking, with AI as a scalable commodity. He announced three strategic pillars: AI models, AI chips, and AI cloud. The Qwen team demonstrated Qwen3.8-Max, a model that autonomously iterated 33 times over a month without human intervention, achieving a score of 45 on the Artificial Analysis benchmark. Alibaba's chip subsidiary Pingtouge unveiled the Zhenwu V900 AI chip, with 3x performance over its predecessor, 216GB memory, and support for clusters up to 500,000 cards, set for mass production in Q1 2027. Separately, DeepSeek appointed 34-year-old former Hillhouse Capital partner Yan Wentao as CFO, ending a three-year vacancy. DeepSeek is nearing the close of its second funding round at a pre-money valuation of 500 billion yuan and is preparing for a potential IPO on Shanghai's STAR Market.
Alibaba Cloud's AIDC Expansion and New AI Chip Usher in New Era of Computing
This research report from Kaiyuan Securities analyzes key announcements from Alibaba's 2026 Apsara Conference. Pingtouge unveiled the Zhenwu V900 AI chip, featuring 216GB memory and 1200GB/s interconnect bandwidth, supporting trillion-parameter model training, with mass production planned for Q1 2027. A new Panjiu supernode server was also released, integrating the V900 chip and enabling AI clusters of up to 500,000 cards. Alibaba Cloud announced plans to establish data centers in Turkey, Finland, and the Netherlands within 12 months, and expand existing sites in Malaysia, Germany, France, UAE, and Hong Kong, targeting a global data center capacity exceeding 20GW by 2032. The report also highlights the Qwen-Audio-3.1 voice model series with price cuts up to 95%, and progress on the Qwen model's recursive self-improvement (RSI) capabilities. The report provides a list of recommended and beneficiary stocks across the AI data center supply chain, including switch chips, optical modules, liquid cooling, and server power supplies.
Alibaba's Pingtouge Unveils Zhenwu V900 AI Chip, Claims 3x Performance, Targets 2027 Mass Production
At the 2026 Hangzhou Yunqi Conference, Alibaba's chip unit Pingtouge (平头哥) announced the Zhenwu V900, a new training-inference integrated AI chip. The company claims the V900 offers three times the single-chip performance of its predecessor, the Zhenwu M890, and is designed to handle trillion-parameter large language models. The chip features 216GB of high-bandwidth memory and 1200GB/s inter-chip interconnect bandwidth. Originally slated for Q3 2027, mass production has been accelerated to Q1 2027, a move Pingtouge Vice President Gao Hui attributed to surging domestic demand for computing power and Alibaba Cloud's engineering capabilities. The article positions the V900 not as a standalone product but as the centerpiece of a broader system-level strategy, including the ICN Switch interconnect chip, the Panmai smart NIC, and the Zhenyue SSD controller, all designed for large-scale cluster deployment up to 500,000 cards. Pingtouge also revealed a roadmap for its Yitian server CPUs (720, 730, 750) to improve CPU-AI chip synergy for agent-based AI workloads. The company's SAIL software stack is highlighted as critical for customer migration and ecosystem adoption. The chip has already served over 650 enterprise clients across various industries.
Show 17 older updatesHide older updates
AI Application Catalysts Accelerate Computing Network Buildout, ChiNext Computing ETF Gains Attention
On September 24, the ChiNext Computing Index fell 2.33% as of 14:28, while individual stock Xieshu Data rose. The ChiNext Computing ETF (E Fund) saw net inflows the previous day. On September 23, the AI application sector received dense catalysts, with industry sources noting that the focus of computing power is shifting from training to inference and edge hardware. On September 22, Alibaba CEO Wu Yongming stated at the Apsara Conference that customer AI demand is very strong and Alibaba Cloud will fully invest in AI infrastructure, targeting over 20GW of global data center capacity by 2032. On the same day, Pingtouge released its strongest domestic AI chip, the Zhenwu V900, with computing power triple that of its predecessor, and Haiguang Information launched its first embedded CPU for industrial control. On September 21, Caitong Shares disclosed a computing power technology service agreement with a leading company totaling at least 1.812 billion yuan. On September 18, a State Council executive meeting studied computing network construction, calling for improved infrastructure. Changjiang Securities noted that cloud vendors at home and abroad continue expansion, computing supply remains tight, and rental price recovery and large orders confirm high industry prosperity.
Read sourceChina's MIIT Plans Compute Expansion; Alibaba, Huawei Launch New AI Chips
On September 24, China's Ministry of Industry and Information Technology (MIIT) announced plans for the '15th Five-Year Plan' period to systematically advance next-generation communication networks, expand and upgrade computing infrastructure, and strengthen computing network coordination to support an integrated computing power network. On September 23, at the 2026 Hangzhou Yunqi Conference, Alibaba revealed progress in AI models, chips, and cloud computing. Its chip subsidiary Pingtouge released the Zhenwu V900, claimed to be the most powerful domestic AI chip, with computing power three times that of its predecessor, scalable to 500,000 cards in a single cluster. Alibaba also stated it will fully invest in AI infrastructure, aiming for over 20GW of global data center capacity by 2032. Separately, on September 22, Huawei's Connect conference unveiled the Ascend 960 supernode, with development ahead of schedule, potentially accelerating domestic AI chip deployment. Changjiang Securities noted that from July to September 2026, global cloud vendors reported capital expenditure and AI business data, with computing power supply remaining tight and rental prices rising, confirming continued industry demand.
Read sourceAlibaba Pushes Full-Stack AI Strategy with $380B Investment, New Chip, and Leadership Changes
At the 2026 Apsara Conference in Hangzhou, Alibaba unveiled its full-stack AI strategy, aiming to become the global 'water, electricity, and coal' of the AI era. CEO Wu Yongming defined the three pillars of the 'machine intelligence' era as AI models, AI chips, and AI cloud. Alibaba announced a target of over 20GW global data center capacity by 2032. Citi raised its 2027-2029 capex forecasts for Alibaba to 2580, 2830, and 2820 billion yuan respectively, assuming partnership cost-sharing. Goldman Sachs estimated annual capex of 2000-3000 billion yuan, calling it manageable via self-built assets, light-asset partnerships, and leasing. Alibaba's AI chip subsidiary Pingtouge released the Zhenwu V900 training-inference chip, set for mass production in Q1 2027. Organizational changes included Liu Daheng taking over the Qwen LLM project and Chen Yusen leading DingTalk's new 'Qwen Office' AI product. Alibaba Cloud's external revenue grew 45% year-on-year, the fastest in 22 quarters, with AI-related product revenue growing triple digits for the 12th consecutive quarter. The article notes 2026 is seen as the first year of AI entering real production environments.
Alibaba Unveils Strongest Domestic AI Chip; Ministry Explores Compute Investment Mechanism
On September 24, Alibaba's chip subsidiary Pingtouge released the Zhenwu V900, described as the most powerful domestically developed AI chip in China, with compute performance three times that of its predecessor, the Zhenwu M890, and a single cluster expandable to 500,000 cards. At the 2026 Hangzhou Yunqi Conference, Alibaba also announced that its next-generation Qwen4 large model is in training, with future Qwen4.5 and Qwen5 versions expected to have 5-10 trillion parameters. The company stated it will fully invest in AI infrastructure, targeting over 20GW of global data center capacity operated by Alibaba Cloud by 2032. Separately, on September 21, China's Ministry of Industry and Information Technology (MIIT) proposed exploring new support mechanisms including compute investment, data investment, and incubation-investment linkage in its AI SME support plan for 2026-2028. Huayuan Securities noted that Alibaba, as a leading cloud provider, is expected to enhance its full-stack capabilities from chips to models and applications, driving domestic AI innovation and sustained demand for domestic computing power. The article also notes that the ChiNext Compute ETF (158050) saw net inflows the previous trading day, despite a 1.94% decline in the ChiNext compute index on September 24.
Read sourceAlibaba Gets Bullish Calls from Multiple Banks After Yunqi Conference AI Pledges
At the 2026 Yunqi Conference in Hangzhou, Alibaba Group CEO Wu Yongming and Qwen LLM project lead Liu Dayiheng outlined the company's long-term AI strategy. Wu pledged full investment in AI infrastructure, targeting global data center capacity exceeding 20GW by 2032. Alibaba's chip subsidiary Pingtouge released the Zhenwu V900, claiming triple the computing power of its predecessor. The company also unveiled the Qwen Book AI agent computer. Following the event, multiple investment banks issued bullish reports. Bank of America maintained a 'Buy' rating with a $175 ADR target. Nomura reiterated 'Buy' with a $178 target, estimating potential external AI cloud revenue of $168 billion by fiscal 2032 if the 20GW capacity target is met. Morgan Stanley maintained 'Overweight' with a $180 target, calling Alibaba a top pick. Huayuan Securities and BOCI also expressed positive views, with BOCI favoring Alibaba over Tencent in the near term due to cloud growth momentum.
Read sourceBofA Securities Reiterates Buy on Alibaba, Citing Cloud Conference and Full-Stack AI Opportunity
BofA Securities released a research report maintaining a 'Buy' rating on Alibaba (HK9988, BABA.US) with a USD 175 price target for US shares and HKD 172 for Hong Kong shares. The report highlights Alibaba's flagship Cloud Computing Conference, where CEO Wu Yongming reiterated a long-term commitment to AI, including investment in AI models, chips, and cloud infrastructure. BofA notes strong and accelerating customer demand, a clear strategic commitment to AI agents, and a self-reinforcing full-stack AI ecosystem spanning infrastructure, models, platforms, and applications. The bank sees Alibaba capturing value across multiple AI stack layers, with near-term monetization from cloud and AI services and longer-term growth from applications and international expansion. A key catalyst is the quarterly update for the period ending September, expected before the October quiet period. Alibaba plans to operate over 20 gigawatts of global data center capacity by 2032, implying annual additions of about 2 GW. BofA estimates construction costs at USD 30-35 billion per GW but expects most to be financed through leases and partnerships, limiting direct capex. The bank maintains its FY2027-2028 capex forecast at RMB 230-250 billion. Alibaba's chip unit, Pingtouge, launched the Zhenwu V900 AI accelerator, described as China's highest-performance AI chip, with performance three times that of the Zhenwu M890 and scalable to 500,000 chips in a single cluster. Alibaba targets increasing self-developed chips' share of deployed computing power from under 10% to 40-50% over time. The Tongyi Qianwen roadmap includes the Qianwen 4.5/5 models with 5-10 trillion parameters, and the upcoming Qianwen 4 model with upgraded multimodal capabilities.
Read sourceBofA Securities Maintains Buy on Alibaba, Citing Full-Stack AI Opportunity at Cloud Summit
BofA Securities reiterated a 'Buy' rating on Alibaba (09988, BABA.US) with a USD 175 price target for US-listed shares and HKD 172 for Hong Kong-listed shares, following Alibaba Cloud's flagship Apsara Conference. CEO Eddie Wu reaffirmed a long-term commitment to AI, emphasizing continued investment in AI models, chips, and cloud infrastructure. The bank observed strong and rapidly accelerating customer demand, a clear strategic commitment to AI agents, and the development of a recursive self-improvement (RSI) architecture. BofA believes Alibaba's industry-leading full-stack AI ecosystem allows it to capture value across multiple layers of the AI stack, with cloud driving near-term monetization and AI services, while applications and international expansion support longer-term growth. A potential catalyst is the September quarter update before the October quiet period. Alibaba plans to operate over 20 gigawatts of global data center capacity by 2032, implying average annual additions of about 2 GW. BofA estimates construction costs at USD 300-350 billion per GW but expects most to be financed through leases and other arrangements, limiting direct capex burden. The bank maintains its FY2027-2028 capex forecast at RMB 230-250 billion. Alibaba's chip unit, Pingtouge, released the Zhenwu V900 AI accelerator, described as China's highest-performing AI chip, with performance three times that of the Zhenwu M890 and scalable to clusters of up to 500,000 chips. Alibaba aims to increase the share of self-developed chips in deployed computing power from under 10% to 40-50% over time. The Tongyi Qianwen roadmap includes plans for the Qianwen 4.5/5 series with 5-10 trillion parameters, and the upcoming Qianwen 4 model with upgraded multimodal capabilities.
Alibaba Cloud targets 20GW global data center capacity by 2032; AI chip and model plans unveiled
Alibaba CEO Wu Yongming announced at the Yunqi Conference that the company will fully invest in AI infrastructure, targeting Alibaba Cloud's global data center capacity to exceed 20GW by 2032. Separately, Alibaba plans to train a new-generation AI model with 5 trillion to 10 trillion parameters. Its chip subsidiary Pingtouge released the new AI chip Zhenwu V900, claiming 3x the computing power of its predecessor M890, with significantly increased annual shipments expected. Alibaba also launched the mobile AI solution Qwen Intelligence and enterprise AI Agent products. On the market front, the ChiNext Computing ETF (158061) saw net inflows of 2.1719 million yuan in the last trading day, with its underlying index dipping 0.39% intraday. The fund's latest net asset value is 370 million yuan. Analysts at Industrial Securities maintain that the computing sector's high prosperity continues, with hardware logic strengthening and the ChiNext Computing Index achieving a historical annualized return of 55.18%.
Read sourceAlibaba unveils AI chip, 20GW data center plan, personal agent at Yunqi Conference
At the 2026 Hangzhou Yunqi Conference, Alibaba Group CEO Eddie Wu outlined an aggressive AI infrastructure strategy, comparing the shift to machine intelligence to the Industrial Revolution. Alibaba targets expanding its global data center capacity to over 20GW by 2032 and plans to train a new AI model with 5 to 10 trillion parameters. Its chip unit, Pingtouge, launched the Zhenwu V900 AI chip, claiming three times the performance of its predecessor and scalability to 500,000 cards per cluster. Wu stated that long-term AI demand far exceeds current supply, citing global supply chain constraints. Separately, Alibaba's Qwen unit announced acceleration of its Personal Agent, aiming to provide personalized AI assistants for 300 million users, with applications in finance, health, and fitness. Following the announcements, Alibaba's Hong Kong-listed shares rose over 4% to a monthly high.
Read sourceAlibaba CEO: AI intelligence becoming scalable utility, plans 5-10T parameter model and new chip
Alibaba Group CEO Eddie Wu stated that intelligence is becoming a scalable utility, with machine-generated thinking currently below 3% of total human cognitive volume but on a path toward 1,000x human capacity. The company's AI roadmap spans three layers: models, chips, and cloud. Its Qwen model is pursuing Recursive Self-Improvement and plans a 5-10 trillion parameter model for complex, longer-horizon tasks on the road to Artificial Superintelligence (ASI). Alibaba also unveiled the Zhenwu V900, described as China's most powerful AI chip, offering 3x the performance of the M890, 216GB memory, 1,200GB/s inter-chip bandwidth, and support for clusters of up to 500,000 cards. Mass production and commercial release are planned for Q1 2027, with expectations for sharply higher annual chip shipments. Additionally, Alibaba Cloud plans to exceed 20GW of global data-center capacity by 2032, as AI demand continues to outpace infrastructure supply.
Read sourceAlibaba's Zhenwu V900 AI Chip Triples Performance, 500,000-Card Cluster Targets 20GW Compute
At the 2026 Apsara Conference, Alibaba Group CEO Wu Yongming emphasized chips as the 'generators' of the AI era. Alibaba's chip design unit, Pingtouge, unveiled the Zhenwu V900 AI training-inference chip, claiming triple the performance of its predecessor, the M890, with 216GB of memory and 1200GB/s interconnect bandwidth. Mass production is slated for Q1 2027. The new Panjiu supernode server integrates the V900 with other chips, enabling a single cluster to scale to 500,000 cards. On the model side, Alibaba announced the training of Qwen4, targeting 5 to 10 trillion parameters, and detailed progress on Recursive Self-Improvement (RSI) for Qwen3.8-Max. The company also outlined a CPU roadmap through 2029. Alibaba Cloud reported Q2 2026 AI cloud revenue of approximately 48.4 billion yuan, up 45% year-on-year, with capital expenditure of 67.7 billion yuan. The long-term goal is to operate over 20GW of global data center capacity by 2032.
Read sourceAlibaba Unveils Strongest Chinese AI Chip, Plans 10-Trillion-Parameter Model, 20GW Data Centers by 2032
At the 2026 Hangzhou Apsara Conference, Alibaba announced a series of major AI infrastructure developments. Its chip unit, Pingtouge, released the Zhenwu V900 AI chip, claiming it is the most powerful domestically developed AI chip in China, with three times the performance of its predecessor, the M890. The chip is slated for mass production and sale in Q1 2027. Alibaba CEO Wu Yongming stated the company aims to train a new model with 5-10 trillion parameters, advancing toward artificial superintelligence (ASI) through recursive self-improvement (RSI). He also set a target for Alibaba Cloud to operate over 20 gigawatts of global data center capacity by 2032. An analyst estimate cited in the report suggests this could translate to approximately 1.2 trillion yuan in external revenue by that year. The announcements also included new server nodes, CPU plans, and the integration of Alibaba's AI phone solution into upcoming Honor smartphones.
Read sourceAlibaba CEO Wu Yongming Sets 20GW Data Center Target by 2032 in AI Infrastructure Push
Alibaba CEO Wu Yongming announced a new AI infrastructure target on September 22, aiming for Alibaba Cloud to operate over 20 gigawatts (GW) of global data center capacity by 2032, comparable to the Three Gorges Dam's 22.5 GW capacity. This goal is part of a broader strategy to position Alibaba as a 'super power plant' for the AI era. Wu stated that the company plans to train a new model with 5-10 trillion parameters, moving toward artificial superintelligence (ASI). He also revealed the next-generation AI chip, Zhenwu V900, with three times the computing power of its predecessor, and the M890 AI supernode supporting efficient inference for models over 2 trillion parameters. The announcement follows Alibaba's February 2025 pledge of 380 billion yuan in AI infrastructure investment over three years. The company's capital expenditure reached 67.68 billion yuan in the June quarter, contributing to a negative free cash flow of 44.67 billion yuan. Analysts cited in the report note that AI-related revenue now accounts for 35% of Alibaba Cloud's external commercial revenue, driven by strong demand for AI inference services that outpaces supply.
Read sourceAlibaba's Pingtouge Releases Strongest Domestic AI Chip; Related Stocks Surge
On September 22, AI chip concept stocks showed strong performance in the Chinese market, with several companies hitting daily price limits. The rally followed news from the 2026 Yunqi Conference, where Alibaba's semiconductor unit Pingtouge unveiled its latest domestic AI chip, the Zhenwu V900. According to a Caixin report, the new chip offers three times the computing power of its predecessor, the Zhenwu M890. Alibaba stated that due to product line maturity and broad customer adoption, annual shipments are expected to increase significantly. The article notes that global tech giants including OpenAI, Nvidia, Meta, Google, and Microsoft are actively developing their own AI chips, while domestic players like Huawei, Cambricon, Baidu, and Moore Threads are also investing heavily. Nvidia CEO Jensen Huang predicted at an AI summit in Scotland that Nvidia chip sales could double by 2027 as AI adoption accelerates across healthcare, manufacturing, and financial services. IDC forecasts the global AI chip market will grow from approximately $48 billion in 2025 to $280 billion by 2030. YiOu think tank projects China's AI chip market will reach 250 billion yuan in 2026 and 1.34 trillion yuan by 2029, representing a compound annual growth rate of 75%.
Read sourceAlibaba Releases Strongest Chinese AI Chip, Hong Kong Stock Rises for Third Day
On September 22, at the 2026 Hangzhou Yunqi Conference, Alibaba's chip unit Pingtouge released the new AI chip Zhenwu V900, which it claims is the most powerful AI chip in China, with three times the computing power of its predecessor. The company stated that a single cluster based on this chip can scale to 500,000 cards, supporting cutting-edge model training and inference. Alibaba expects a significant increase in annual shipments of Pingtouge AI chips. Following the news, Alibaba's Hong Kong-listed stock (9988.HK) rose over 4% intraday and closed nearly 2% higher, marking a third consecutive day of gains with a cumulative increase of nearly 10%. Separately, Alibaba CEO Wu Yongming delivered a speech at the conference, outlining his vision for the 'machine intelligence' era. He argued that machines will become the primary thinking force, handling 99.9% of thought work, and that AI models, AI chips, and AI cloud are the three pillars of this era. He detailed Alibaba's plans: the Qwen team is exploring recursive self-improvement (RSI) towards artificial superintelligence (ASI) and plans to train a 5-10 trillion parameter model; Pingtouge is building a full lineup of data center chips including GPU, CPU, smart NICs, and interconnect chips; and Alibaba Cloud aims to operate over 20GW of global data center capacity by 2032 to meet surging AI demand.
Read sourceAlibaba Unveils New AI Chip, Targets 20GW Data Center Capacity by 2032
Alibaba (BABA.N) announced its next-generation self-developed AI accelerator chip, the Zhenwu V900, at the Apsara Conference on September 22. The chip, designed by Alibaba's T-Head Semiconductor, offers approximately three times the performance of its predecessor and can form clusters of up to 500,000 chips for advanced large model training. CEO Wu Yongming stated that Alibaba expects exponential growth in AI demand, necessitating expanded underlying computing power. The company also set a target to expand Alibaba Cloud's global data center capacity to 20GW by 2032, a tenfold increase from 2022 levels. This is part of a previously announced 380 billion yuan ($53 billion) investment in AI and cloud infrastructure over three years. The article notes that Alibaba's strategy is shifting from single-chip performance to ultra-large-scale clusters, competing with Nvidia's ecosystem. The 20GW target quantifies future capital expenditure on power, cooling, and networking. Alibaba reported AI cloud revenue of $7.1 billion, up 45% year-on-year, the fastest growth in 22 quarters, supporting its investment thesis. The article highlights that unlike US tech giants, Alibaba must also address restrictions on high-end chip access, making self-developed semiconductors critical to its expansion plans.
Read sourceAlibaba Cloud Targets 20GW Data Center Capacity by 2032, T-Head Unveils AI Chip Zhenwu V900
At the 2024 Apsara Conference, Alibaba Group CEO Wu Yongming announced that Alibaba Cloud aims to operate over 20GW of global data center capacity by 2032. Concurrently, Alibaba's chip unit T-Head released the Zhenwu V900, touted as the most powerful domestic AI chip, which could reduce Alibaba Cloud's computing costs. Analysts estimate that at $12-15 billion annual revenue per GW, this expansion could enable Alibaba Cloud to achieve its 2030 external revenue target of $100 billion earlier, potentially exceeding one trillion yuan by 2032. The article also notes two major domestic AI chip financing rounds: RISC-V cloud AI chip firm Yixin Intelligent raised nearly 2 billion yuan (valuation ~15 billion yuan), and GPGPU company Xingyun Jicheng raised nearly 800 million yuan led by CATL's investment platform. Separately, the People's Bank of China held a symposium with foreign financial institutions, and the yuan's central parity rate hit a three-year high against the USD, signaling independent appreciation driven by capital inflows and policy communication.
Read source