CloudTech is part of the TechForge Publications seriesView AllAI NewsDeveloperIoT NewsMarketing TechTechHQTech Wire AsiaTelecomsView AllAI NewsDeveloperIoT NewsMarketing TechTechHQTech Wire AsiaTelecomsTechForge SearchCategoriesCloud in ActionCloud MigrationCloud ROI & CostInternal Change ManagementMissteps & LessonsSME & Startup CloudEditorial DeskAnnouncements & AnalysisForecasts & TrendsMigrations: Behind the ScenesTechEx EventsFeaturesInterviewsPodcastsSponsored ContentVideosWebinarsFuture of CloudAI & CloudCloud EthicsEdge & Distributed CloudOpen CloudQuantum & CloudServerless ArchitectureSustainable CloudIndustry PerspectivesEducation & ResearchFinanceHealthcare & Life SciencesLegal & HRMedia, Gaming & CreativePublic SectorRetail & ConsumerMarket IntelligenceCloud StartupsEarnings & Market ShareEvent CoverageMergers & AcquisitionsVendor Roadmaps & LeadershipSecurity, Privacy & TrustCloud CybersecurityCyber Security & Cloud ExpoEncryption & Data PrivacyGovernance, Risk & ComplianceIdentity & AccessStrategy & Decision-MakingChoosing a Cloud StrategyFinOps & BudgetsLock-In & ExitMulti- & Hybrid CloudProcurement & ContractsSkills & HiringTechnology StackBig VendorsContainers & KubernetesDatabases & Data PlatformsInfrastructure as CodeObservability & MonitoringXaaS ModelsEventsResourcesVideo InterviewsPodcastsAll ResourcesMoreEditorial OpportunitiesAbout UsContact Us SearchCategoriesCloud in ActionCloud MigrationCloud ROI & CostInternal Change ManagementMissteps & LessonsSME & Startup CloudEditorial DeskAnnouncements & AnalysisForecasts & TrendsMigrations: Behind the ScenesTechEx EventsFeaturesInterviewsPodcastsSponsored ContentVideosWebinarsFuture of CloudAI & CloudCloud EthicsEdge & Distributed CloudOpen CloudQuantum & CloudServerless ArchitectureSustainable CloudIndustry PerspectivesEducation & ResearchFinanceHealthcare & Life SciencesLegal & HRMedia, Gaming & CreativePublic SectorRetail & ConsumerMarket IntelligenceCloud StartupsEarnings & Market ShareEvent CoverageMergers & AcquisitionsVendor Roadmaps & LeadershipSecurity, Privacy & TrustCloud CybersecurityCyber Security & Cloud ExpoEncryption & Data PrivacyGovernance, Risk & ComplianceIdentity & AccessStrategy & Decision-MakingChoosing a Cloud StrategyFinOps & BudgetsLock-In & ExitMulti- & Hybrid CloudProcurement & ContractsSkills & HiringTechnology StackBig VendorsContainers & KubernetesDatabases & Data PlatformsInfrastructure as CodeObservability & MonitoringXaaS ModelsEventsResourcesVideo InterviewsPodcastsAll ResourcesMoreEditorial OpportunitiesAbout UsContact Us Subscribe Subscribe SearchCategoriesCloud in ActionCloud MigrationCloud ROI & CostInternal Change ManagementMissteps & LessonsSME & Startup CloudEditorial DeskAnnouncements & AnalysisForecasts & TrendsMigrations: Behind the ScenesTechEx EventsFeaturesInterviewsPodcastsSponsored ContentVideosWebinarsFuture of CloudAI & CloudCloud EthicsEdge & Distributed CloudOpen CloudQuantum & CloudServerless ArchitectureSustainable CloudIndustry PerspectivesEducation & ResearchFinanceHealthcare & Life SciencesLegal & HRMedia, Gaming & CreativePublic SectorRetail & ConsumerMarket IntelligenceCloud StartupsEarnings & Market ShareEvent CoverageMergers & AcquisitionsVendor Roadmaps & LeadershipSecurity, Privacy & TrustCloud CybersecurityCyber Security & Cloud ExpoEncryption & Data PrivacyGovernance, Risk & ComplianceIdentity & AccessStrategy & Decision-MakingChoosing a Cloud StrategyFinOps & BudgetsLock-In & ExitMulti- & Hybrid CloudProcurement & ContractsSkills & HiringTechnology StackBig VendorsContainers & KubernetesDatabases & Data PlatformsInfrastructure as CodeObservability & MonitoringXaaS ModelsEventsResourcesVideo InterviewsPodcastsAll ResourcesMoreEditorial OpportunitiesAbout UsContact Us Hamburger Toggle Menu AI & Cloud, Announcements & Analysis, Big Vendors, Cloud Computing, Cloud in Action, Features, Vendor Roadmaps & LeadershipAlibaba Cloud plans 20GW AI infrastructure expansion by 2032Muhammad Zulhusni23rd September 2026 Share this story: Tags:AIai infrastructureAlibaba Cloudcloudcloud infrastructuredata centreCategories::AI & CloudAnnouncements & AnalysisBig VendorsCloud ComputingCloud in ActionFeaturesVendor Roadmaps & LeadershipAlibaba Cloud plans to expand its global data centre capacity to more than 20 gigawatts by 2032 as the company increases infrastructure spending for AI computing.Alibaba Group CEO Eddie Wu announced the target at the company’s Apsara Conference in Hangzhou yesterday.Alibaba did not disclose its current data centre capacity or provide a regional breakdown for the planned expansion.The target follows Alibaba’s February 2025 commitment to invest more than 380 billion yuan, or about US$52 billion at the time, in cloud computing and AI infrastructure over three years.The company said the spending would exceed its investment in those areas over the previous decade.The expansion covers more than data centre capacity.
Alibaba is also developing its own AI accelerators, server processors, networking and storage systems, cloud services, and Qwen models.Wu described the strategy at Apsara as covering the full AI technology stack, including models, semiconductors, and data centres.In the June quarter of 2026, Alibaba’s AI Cloud and Compute Services revenue reached US$7.1 billion, up 45% year over year.Adjusted earnings before interest, taxes, and amortisation for the cloud segment rose 133% to US$830 million, while its adjusted EBITA margin reached about 12%.
Alibaba said revenue from AI-related products recorded triple-digit growth for the twelfth consecutive quarter.Reuters reported in August that Alibaba had already spent about half of its planned 380 billion yuan AI investment during 2026 as the company increased spending on proprietary chips and AI infrastructure.Wu also said Alibaba expected margins to improve as it deployed more processors developed by its T-Head semiconductor unit and replaced some commercially procured chips in its data centres.Wu said at Apsara that customer demand for AI services was exceeding Alibaba Cloud’s available supply, according to .He added that supply-chain constraints were limiting how quickly the company could expand capacity and said Alibaba Cloud would begin bringing AI supernodes into commercial operation at scale during the current quarter.Alibaba builds more of its AI infrastructureAlibaba also introduced the Zhenwu V900, an AI processor designed by its T-Head semiconductor unit for model training and inference.
The processor has 216GB of memory and 1,200GB/s of inter-chip bandwidth, with support for data formats including FP8 and FP4.The company said the V900 delivers three times the performance of the Zhenwu M890, which Alibaba introduced in May.The V900 is scheduled to enter mass production and commercial release in the first quarter of 2027.The V900 extends the processor roadmap Alibaba outlined with the M890 earlier this year.Reuters reported that Alibaba developed the Zhenwu line as Chinese technology companies sought domestic alternatives to Nvidia processors amid US restrictions on advanced AI chip exports to China.Alibaba’s earlier M890 launch also connected its custom processors more closely with its cloud infrastructure.
The company introduced the Panjiu AL128 server system with 128 M890 accelerators in a rack and made it available to enterprise customers through Alibaba Cloud’s Bailian platform.T-Head said in May that it had shipped more than 560,000 Zhenwu processors to more than 400 external customers across 20 industries.Alibaba is integrating the V900 into an upgraded supernode server alongside its own networking, storage, and interface components.The system is designed to support clusters containing as many as 500,000 accelerator cards, according to the company.Alibaba’s chip roadmap also covers general-purpose processors.
The company plans to introduce its Yitian 720 and Yitian 730 CPUs in 2027, with the latter based on T-Head’s own microarchitecture.Alibaba has already deployed T-Head processors within its cloud infrastructure.In a May shareholder letter, the company said its proprietary AI chips had entered production at scale and were supplying computing capacity to Alibaba Cloud infrastructure and its model-as-a-service inference platform.Alibaba Cloud also announced updates to the networking and storage systems used by large AI clusters.Its HPN 8.0 Pro networking architecture provides 100 petabits of bandwidth and supports more than 130,000 network ports operating at 800Gbps within a single cluster, according to Alibaba.Its Cloud Parallel File Storage system is designed for AI training workloads and supports throughput measured in hundreds of terabytes per second, as well as hundreds of millions of input/output operations per second.
Alibaba said the system can reduce AI storage costs by 69%, although the company did not provide independent benchmark data alongside the announcement.As accelerator clusters grow, networking bandwidth and storage throughput become more important to keeping compute resources utilised.Google has identified similar constraints in its own AI infrastructure, particularly as larger clusters place greater demands on networks and storage systems.Google’s AI Hypercomputer architecture combines processors, networking, storage, and orchestration software.Its infrastructure includes Google TPUs and Axion CPUs alongside Nvidia GPUs and Intel and AMD processors, as well as its Virgo data centre network and high-performance storage systems.Hyperscalers develop more custom infrastructureAWS, Microsoft, and Google have also developed custom processors while continuing to use hardware from outside suppliers.
AWS operates Trainium AI processors and Graviton CPUs, Microsoft uses Maia accelerators and Cobalt CPUs across Azure, while Google combines its TPUs and Axion CPUs with Nvidia, Intel, and AMD hardware.Amazon has cited price-performance and infrastructure economics as reasons for developing its own processors.The company said Trainium2 offered about 30% better price-performance than comparable GPUs, while Trainium3 improved price-performance by another 30% to 40% over Trainium2.Those figures are based on Amazon’s own measurements.AWS also uses custom silicon for general-purpose cloud workloads.
Amazon introduced Graviton in 2018 and said in April 2026 that more than 90,000 customers were using Graviton-based infrastructure.Microsoft said in its fiscal 2026 third-quarter earnings call that millions of servers across its fleet use Microsoft-designed networking, security, and virtualisation silicon.Its Maia 200 AI accelerator was operating in data centres in Iowa and Arizona, while Cobalt CPUs had been deployed in nearly half of Microsoft’s data centre regions.All three providers continue to use third-party hardware alongside their own silicon.Microsoft deploys its processors alongside Nvidia and AMD hardware, Google supports TPUs alongside Nvidia GPUs, and AWS has said it will continue offering Nvidia systems as it expands Trainium.Alibaba connects infrastructure to cloud servicesThe hardware sits underneath Alibaba’s cloud services for deploying and operating AI applications.
Its AgentCore platform is designed to manage agents through development and deployment, while Agent Security Center provides security and compliance controls for agent-based applications.Alibaba said in a May shareholder letter that the growth focus of its cloud business was moving from traditional compute and storage toward models, AI computing, and agent services.Alibaba’s infrastructure plans also extend to larger Qwen models.Qwen 4 is currently in training, while future Qwen 4.5 and Qwen 5 models are planned with between five trillion and 10 trillion parameters, compared with 2.4 trillion parameters for the current Qwen 3.8 Max model.See also: Tencent and Alibaba expand APAC cloud infrastructureWant to learn more about Cloud Computing from industry leaders? Check out Cyber Security & Cloud Expo taking place in Amsterdam, California, and London.The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.Cloud Computing News is powered by TechForge Media.
Explore other upcoming enterprise technology events and webinars here.About the Author Muhammad ZulhusniJournalist As a tech journalist, Zul focuses on topics including cloud computing, cybersecurity, and disruptive technology in the enterprise industry.He has expertise in moderating webinars and presenting content on video, in addition to having a background in networking technology.Related Fully Managed vs.WPCO: which cloud-based Android management mode is right for your company?23rd September 2026 AWS AgentCore prompt injection exposes credential risks21st September 2026 AWS cloud data loss in Bahrain and UAE raises resilience questions18th September 2026 Why cloud resilience is failing in the modern era17th September 2026 Fully Managed vs.
WPCO: which cloud-based Android management mode is right for your company?23rd September 2026 AWS AgentCore prompt injection exposes credential risks21st September 2026 AWS cloud data loss in Bahrain and UAE raises resilience questions18th September 2026 Why cloud resilience is failing in the modern era17th September 2026 Join our CommunitySubscribe now to get all our premium content and latest tech news delivered straight to your inbox Click here Popular XaaS ModelsSoftLayer beefs up its bare metal offering, available on hourly basis 44573 view(s)Cloud ROI & Cost, Interviews, Sponsored Content, Sustainable CloudRipple effect: Xylem’s sustainable water solutions for Europe’s data centres 20748 view(s)Cloud Computing, XaaS ModelsConcern over cloud storage security remains says Spiceworks – but good news for OneDrive 12759 view(s)Big Vendors, Cloud Computing, Cloud Cybersecurity, Market Intelligence, Security, Privacy & Trust10 real-life cloud security failures and what we can learn from them 7275 view(s)XaaS ModelsSoftLayer beefs up its bare metal offering, available on hourly basis 44573 view(s)Cloud ROI & Cost, Interviews, Sponsored Content, Sustainable CloudRipple effect: Xylem’s sustainable water solutions for Europe’s data centres 20748 view(s)Cloud Computing, XaaS ModelsConcern over cloud storage security remains says Spiceworks – but good news for OneDrive 12759 view(s)Big Vendors, Cloud Computing, Cloud Cybersecurity, Market Intelligence, Security, Privacy & Trust10 real-life cloud security failures and what we can learn from them 7275 view(s) See all Latest View All Latest Cloud in Action18th September 2026AWS cloud data loss in Bahrain and UAE raises resilience questions Cloud Computing17th September 2026Why cloud resilience is failing in the modern era Features14th September 2026UAE redesigns 5GW AI data centre plan around physical security Cloud in Action18th September 2026AWS cloud data loss in Bahrain and UAE raises resilience questions Cloud Computing17th September 2026Why cloud resilience is failing in the modern era Features14th September 2026UAE redesigns 5GW AI data centre plan around physical security SubscribeAll our premium content and latest tech news delivered straight to your inbox Subscribe ExploreAbout UsContact UsNewsletterPrivacy PolicyCookie PolicyAbout UsContact UsNewsletterPrivacy PolicyCookie PolicyReach Our AudienceAdvertisePost a Press ReleaseContact UsAdvertisePost a Press ReleaseContact UsCategoriesCloud in ActionEditorial DeskFeaturesFuture of CloudIndustry PerspectivesMarket IntelligenceSecurity, Privacy & TrustTechnology StackStrategy & Decision-MakingAll CategoriesCloud in ActionEditorial DeskFeaturesFuture of CloudIndustry PerspectivesMarket IntelligenceSecurity, Privacy & TrustTechnology StackStrategy & Decision-MakingAll CategoriesOther PublicationsExplore AllAI NewsDeveloperIoT NewsMarketing TechTechHQTech Wire AsiaTelecomsExplore AllAI NewsDeveloperIoT NewsMarketing TechTechHQTech Wire AsiaTelecomsCloudTech News is part of TechForge SubscribeAll our premium content and latest tech news delivered straight to your inbox
Read More
