AWS to add 2 million Nvidia GPUs as cloud capacity demand grows

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, Features, Future of Cloud, Market IntelligenceAWS to add 2 million Nvidia GPUs as cloud capacity demand growsMuhammad Zulhusni27th August 2026 Share this story: Tags:AIAWScloudcloud infrastructuredata centreNVIDIACategories::AI & CloudAnnouncements & AnalysisBig VendorsFeaturesFuture of CloudMarket IntelligenceAWS plans to deploy another two million Nvidia GPUs across its global infrastructure in 2027 and 2028, adding to an earlier commitment for more than one million GPUs.The additional deployment will include Nvidia Blackwell Ultra, Rubin, and Rubin Ultra GPUs.The agreement also covers Nvidia CPUs, networking, and interconnect technology that AWS plans to incorporate into its infrastructure.AWS announced its previous commitment at Nvidia GTC in March, when it said it would add more than one million Nvidia GPUs starting in 2026.Nvidia later told Reuters that deliveries would begin in 2026 and continue through the end of 2027, taking AWS’s planned Nvidia deployments across the two commitments to more than three million GPUs.AWS did not provide an update on deployment progress for the earlier commitment in its August 26 announcement.The latest agreement also includes plans for 100,000 Nvidia GPUs on secure AWS infrastructure for US federal and national-security workloads.

The companies said the systems would form part of AI infrastructure being developed for the US government.AWS plans capacity years aheadThe new commitment comes as AWS already has customer reservations for computing capacity extending into 2028.Amazon CEO Andy Jassy said most of the AWS compute capacity being added for 2027 was already reserved, with customer reservations also extending into 2028.AWS ended the second quarter with a contracted backlog of $496 billion, up from $364 billion three months earlier.Revenue at the cloud business reached $42.2 billion during the quarter, an increase of 37% from a year earlier.Jassy said Amazon still expected its available computing capacity to fall short of demand.

“Even at that amount, we will still not have enough capacity to meet all of the demand we have in 2026,” Jassy said.Amazon’s cash capital expenditure reached $53.1 billion during the second quarter of 2026, compared with $31.4 billion a year earlier.For the first six months of 2026, cash capital expenditure reached $96.3 billion, up from $55.6 billion in the same period of 2025.Amazon said the expenditures primarily reflected technology infrastructure investments, the majority of which support AWS growth, as well as additional capacity for its fulfilment network.Amazon raised its 2026 capital expenditure forecast from about $200 billion to $220 billion in July.Jassy said higher memory costs were a primary factor behind the increase.Combined 2026 capital expenditure estimates for Amazon, Microsoft, Alphabet, Meta, and Oracle rose from about $485 billion in January to around $730 billion by July, according to a Reuters analysis of LSEG data.

The figures include spending outside AI because the companies do not consistently disclose AI-specific capital expenditure.AWS’s infrastructure investments also carry long lead times.Amazon said AWS typically commits capital to land, power, buildings, chips, servers, and networking equipment six to 24 months before it begins billing customers for the resulting capacity.Amazon said data centres have useful lives of more than 30 years, compared with five to six years for chips, servers, and networking equipment.Jassy separately said Amazon can begin spending on a data centre roughly two years before it opens.Trainium grows alongside NvidiaAWS is increasing its Nvidia deployments while continuing to expand its own AI processor portfolio.

Trainium3 began shipping at the start of 2026, while Trainium4 is expected to begin deliveries in 2027.Amazon said in May that Trainium3 was nearly fully subscribed and much of its planned Trainium4 capacity had already been reserved.The company also said it had more than $225 billion in revenue commitments for Trainium.Trainium is designed for AI training and inference workloads and provides AWS with its own alternative to third-party accelerators.Amazon has developed the processor through Annapurna Labs, the chip-design company it acquired in 2015.Amazon has linked Trainium to the cost of expanding AWS infrastructure.

In its shareholder letter, the company said it expects Trainium, at scale, to save tens of billions of dollars in annual capital expenditure and provide several hundred basis points of operating-margin advantage compared with relying on other chips for inference.AWS is therefore adding capacity across both its own accelerators and Nvidia hardware.The March and August Nvidia agreements alone cover more than three million planned GPUs.Nvidia technology is also being incorporated into AWS’s custom silicon infrastructure.Trainium4 is being designed to support Nvidia’s NVLink Fusion interconnect technology, which provides high-bandwidth connections between processors and accelerators inside AI systems.The collaboration extends beyond GPUs.

The earlier agreement included Nvidia ConnectX and Spectrum-X networking equipment, while the August agreement includes work to bring Vera CPU-based infrastructure to AWS and expand NVLink Fusion support with Nvidia’s custom high-bandwidth memory technology.Nvidia and Amazon’s Annapurna Labs said the memory and interconnect work is intended for future Trainium-based infrastructure, allowing Trainium accelerators and Nvidia GPUs to operate within a common rack-scale architecture.Nvidia is expanding those deployments while continuing to face component constraints.The company reported $89 billion in data-centre revenue for its fiscal second quarter, more than double the year-earlier figure.“We are supply-constrained,” Nvidia CFO Colette Kress said during the company’s earnings call.Nvidia said higher memory and component costs would also continue to pressure its margins.Nvidia said specialist AI cloud providers including CoreWeave and Nebius were set to end 2026 with more than eight gigawatts of Nvidia GPU capacity, up from three gigawatts at the end of 2025.The Vera Rubin platform has also begun shipping to customers.

Nvidia previously named AWS, Google Cloud, Microsoft, and Oracle Cloud Infrastructure among the first cloud providers expected to deploy Vera Rubin-based instances.Additional GPU capacity also requires investment in power, cooling, and networking infrastructure.The International Energy Agency expects global data-centre electricity consumption to rise from around 415 TWh in 2024 to about 945 TWh in 2030, with accelerated servers accounting for almost half of the projected increase.Electricity consumption from accelerated servers, mainly driven by AI adoption, is projected to increase by about 30% annually under the IEA’s base case.Cooling and other infrastructure account for another 20% of the projected increase in data-centre electricity consumption through 2030.The IEA estimates that data centres can become operational within two to three years, while the supporting energy infrastructure generally requires longer planning and construction periods.See also: Amazon shifts AWS workloads as power constraints tightenWant 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 Azure VMware Solution licensing changes set October BYOL deadline21st August 2026 What You Need to Know About ISO 27017 Certification20th August 2026 Tencent and Alibaba expand APAC cloud infrastructure19th August 2026 Amazon shifts AWS workloads as power constraints tighten17th August 2026 Azure VMware Solution licensing changes set October BYOL deadline21st August 2026 What You Need to Know About ISO 27017 Certification20th August 2026 Tencent and Alibaba expand APAC cloud infrastructure19th August 2026 Amazon shifts AWS workloads as power constraints tighten17th August 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 35113 view(s)Cloud ROI & Cost, Interviews, Sponsored Content, Sustainable CloudRipple effect: Xylem’s sustainable water solutions for Europe’s data centres 20675 view(s)Cloud Computing, XaaS ModelsConcern over cloud storage security remains says Spiceworks – but good news for OneDrive 12711 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 7067 view(s)XaaS ModelsSoftLayer beefs up its bare metal offering, available on hourly basis 35113 view(s)Cloud ROI & Cost, Interviews, Sponsored Content, Sustainable CloudRipple effect: Xylem’s sustainable water solutions for Europe’s data centres 20675 view(s)Cloud Computing, XaaS ModelsConcern over cloud storage security remains says Spiceworks – but good news for OneDrive 12711 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 7067 view(s) See all Latest View All Latest AI & Cloud19th August 2026Tencent and Alibaba expand APAC cloud infrastructure Cloud Computing17th August 2026Amazon shifts AWS workloads as power constraints tighten AI & Cloud13th August 2026IBM Cloud and Together AI expand AI infrastructure with NVIDIA AI & Cloud19th August 2026Tencent and Alibaba expand APAC cloud infrastructure Cloud Computing17th August 2026Amazon shifts AWS workloads as power constraints tighten AI & Cloud13th August 2026IBM Cloud and Together AI expand AI infrastructure with NVIDIA 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