Microsoft said on November 3, 2025, that it plans to invest $15.2 billion in the United Arab Emirates from 2023 through 2029. The announcement combines spending already made with future commitments, and it refers to more than 80,000 A100-equivalent Nvidia chips enabled by U.S. export licenses—not a confirmed shipment of 80,000 identical GPUs already installed and operating.
The figures at a glance
| Item | What Microsoft reported |
|---|---|
| Total UAE investment | $15.2 billion planned across 2023–2029 |
| Spending reported through 2025 | Just over $7.3 billion |
| Planned spending for 2026–2029 | More than $7.9 billion |
| Earlier licensed capacity | 21,500 A100-equivalent GPUs using A100, H100 and H200 chips |
| Later approved capacity | 60,400 additional A100-equivalent chips involving GB300 systems |
| Announced data-center expansion | 200 megawatts through G42 subsidiary Khazna Data Centers |
| Main UAE strategic partner | G42 |
Microsoft’s announcement is available from Microsoft. The UAE government described the combined figure as more than 80,000 A100-class chips, but the underlying Microsoft description uses an equivalence measure.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
ASUS Dual GeForce RTX 5060 Ti 16GB GDDR7 OC Edition Gaming Graphics Card | $794.37 | Buy on Amazon |
| 2 |
|
ASUS TUF Gaming GeForce RTX™ 5080 16GB GDDR7 OC Edition Graphics Card | $1,814.90 | Buy on Amazon |
What the $15.2 billion actually covers
The headline is a cumulative, multi-year spending plan, not a new $15.2 billion cash transfer announced in November 2025. Microsoft said it had spent just over $7.3 billion across 2023–2025 and planned to spend more than $7.9 billion from 2026 through 2029.
Existing and planned categories
- G42 equity investment: Microsoft’s $1.5 billion minority investment announced in April 2024.
- Infrastructure capital expenditure: More than $4.6 billion through 2025 and more than $5.5 billion planned for 2026–2029.
- Local operating costs and cost of goods sold: More than $1.2 billion through 2025, with almost $2.4 billion planned for 2026–2029.
- Workforce, training and research activity: Included in the company’s overall UAE spending description.
Consequently, the genuinely incremental commitment highlighted by the announcement is the more than $7.9 billion planned for 2026–2029. The $15.2 billion total should not be treated as construction spending alone or as money paid directly to Nvidia.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →#1 Best Overall
- AI Performance: 767 AI TOPS
- OC mode: 2632 MHz (OC mode)/ 2602 MHz (Default mode)
- Powered by the NVIDIA Blackwell architecture and DLSS 4
- Axial-tech fan design features a smaller fan hub that facilitates longer blades and a barrier ring that increases downward air pressure
- A 2.5-slot design maximizes compatibility and cooling efficiency for superior performance in small chassis
What “80,000 Nvidia GPUs” leaves out
Microsoft’s wording separates two U.S.-approved tranches. Earlier licenses enabled the equivalent of 21,500 Nvidia A100 GPUs, using a mix of A100, H100 and H200 chips. Licenses approved in September 2025 enabled shipment of an additional 60,400 A100-equivalent chips involving Nvidia GB300 systems.
A100-equivalent is a comparison unit
An A100-equivalent figure expresses aggregate computing capacity against a reference GPU. It does not mean every physical unit is an A100, nor that a newer GB300 system contains one chip that can be counted literally as one A100. The combined headline is therefore more than 80,000 A100-equivalent chips, based on Microsoft’s accounting.
License, shipment and operation are different stages
- License approval: The U.S. Commerce Department permits specified exports subject to conditions.
- Authorized shipment: Hardware can be sent under those permissions.
- Arrival and installation: Equipment reaches an approved facility and is integrated into a cluster.
- Commercial availability: Capacity is allocated to Azure, G42, government or enterprise users.
- Productive use: Customers run models and applications on the operating system, network and power infrastructure.
Microsoft’s announcement establishes licensing and enabled shipments; it does not provide a complete inventory showing the exact model mix, delivery dates, installation status, utilization or customer allocation. It also does not establish that all 60,400 additional equivalent units had arrived in the UAE.
Why U.S. export licenses matter
Advanced AI accelerators are subject to U.S. national-security and technology-transfer controls. Microsoft said it secured licenses from the Commerce Department to ship Nvidia hardware to the UAE under cybersecurity, national-security and other technology safeguards, including updated safeguards associated with the September 2025 approvals.
Recommended Free Tools
The public announcement does not spell out every license condition. Important unanswered questions include:
- Whether particular systems are operated by Microsoft, G42 or both;
- Who controls hardware, administrator accounts and physical access;
- Which customers, models, data and personnel are permitted;
- Whether equipment can be resold, transferred or moved outside approved environments;
- How compliance is audited and reported.
Approval should therefore not be read as unrestricted UAE access to every advanced Nvidia product. The arrangement is also significant for other U.S. partners: it tests whether Washington can permit strategic AI infrastructure in an allied country while maintaining enforceable controls.
Microsoft, G42 and Khazna: who is building what?
G42 is a UAE-based AI and technology holding group, not merely a construction contractor. In April 2024, Microsoft announced a $1.5 billion minority investment, a board seat for Microsoft Vice Chair Brad Smith and cooperation to expand Azure-based AI infrastructure and solutions across the Middle East, Central Asia and Africa. Details are in Microsoft’s partnership announcement.
G42’s subsidiary Khazna Data Centers is the delivery vehicle named for a new 200MW data-center expansion. G42 and Microsoft said capacity is expected to begin coming online before the end of 2026, supporting Azure AI, cloud services and sovereign-cloud capabilities. See G42’s project announcement.
Megawatts measure facility power capacity. They do not directly reveal GPU count, model-training throughput, customer compute-hours, revenue or delivered performance. The announcements also do not fully establish how the 200MW expansion relates to every existing Microsoft or G42 site.
Rank #2
- Powered by the NVIDIA Blackwell architecture and DLSS 4. System Requirements: Minimum 850W PSU with 16-pin 12V-2x6 (12VHPWR) connector required. Verify before purchasing.
- Military-grade components deliver rock-solid power and longer lifespan for ultimate durability. Compatibility: 348mm (13.7") length, 3.6 slots, 4.3 lbs. Confirm case clearance and slot spacing. GPU bracket included.
- Protective PCB coating helps protect against short circuits caused by moisture, dust, or debris
- 3.6-slot design with massive fin array optimized for airflow from three Axial-tech fans
- Phase-change GPU thermal pad helps ensure optimal thermal performance and longevity, outlasting traditional thermal paste for graphics cards under heavy loads
How Stargate UAE fits in—and where it does not
Stargate UAE is a related but separate initiative. In May 2025, G42, OpenAI, Oracle, Nvidia, SoftBank and Cisco announced a proposed AI infrastructure cluster within a broader UAE–U.S. AI campus. OpenAI described a planned 1GW cluster, with 200MW expected to go live in 2026; its announcement is at OpenAI.
Microsoft’s UAE investment overlaps with Stargate through G42 and the UAE’s wider AI build-out, but the $15.2 billion plan should not automatically be labeled Stargate UAE. Microsoft’s Azure/G42 facilities, Khazna data centers, Abu Dhabi sovereign-cloud work and the OpenAI-led Stargate project have different participants and ownership arrangements.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What the infrastructure is intended to support
Microsoft says the planned capacity will provide access to models from OpenAI, Anthropic, open-source providers and Microsoft, as well as Copilot and AI applications from local and international developers. Those are corporate plans, not an independent audit of model availability, performance or launch dates.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Why the UAE is strategically useful
- It is a wealthy market with strong government support for AI and digital infrastructure.
- Its location can reduce latency for customers across the Middle East, Africa and South Asia.
- Government and regulated-industry buyers increasingly seek local data residency and operational controls.
- Arabic and bilingual AI applications create regional demand that global platforms want to serve.
- Close U.S.–UAE technology and investment ties make large infrastructure projects easier to coordinate.
Microsoft cited its own AI Diffusion Report, which put UAE generative-AI use at 59.4% of the population, compared with 58.6% in Singapore. That is a Microsoft-published statistic, not an uncontested global benchmark.
What “sovereign cloud” means in practice
Sovereign-cloud offerings generally address data residency, administrative control, government access rules, operational personnel, security controls, local compliance and limits on cross-border data movement. They do not automatically mean the UAE owns all underlying hardware, nor do they erase foreign legal obligations.
Abu Dhabi separately announced a sovereign-cloud agreement with Microsoft and Core42 in March 2025 under its Government Digital Strategy 2025–2027. The government’s announcement is at dge.gov.ae. That agreement is part of the broader ecosystem but is not proof that every GPU facility in the UAE has identical sovereignty terms.
What customers can—and cannot—assume today
A headline investment or export license does not prove that a public, on-demand GPU service is available immediately. A buyer would need to confirm:
- the exact UAE region, availability zone and facility;
- the GPU model actually offered, rather than the model mentioned in an authorization;
- on-demand versus reserved capacity, quotas and minimum commitments;
- data-residency, administrator-access and cross-border terms;
- model-service availability, networking and storage performance;
- customer-screening and export-control requirements; and
- whether pricing is public or negotiated.
Azure customers can check current regional products and pricing through the Azure pricing calculator and Azure AI services pages. Microsoft’s infrastructure geography is listed at Azure global infrastructure. Those pages, rather than the investment headline, determine whether a specific SKU can be purchased.
What remains unconfirmed
- The exact physical number and model mix of GPUs delivered;
- which sites receive each hardware tranche;
- commercial launch dates and customer pricing;
- allocation among Microsoft, G42, Azure customers and public-sector users;
- detailed export-license conditions and compliance reporting;
- whether all 60,400 additional A100-equivalent units will be deployed in the UAE by a stated date; and
- the final ownership and operating structure of every facility.
Bottom line
Microsoft’s UAE announcement represents a major multi-year infrastructure and strategic partnership commitment. The accurate reading is: $15.2 billion of cumulative UAE spending for 2023–2029, more than $7.9 billion of planned 2026–2029 spending, licensed capacity equivalent to more than 80,000 A100s, and a separate 200MW G42/Khazna expansion. None of those figures, by themselves, proves that 80,000 identical Nvidia GPUs are already operating or available to every UAE customer.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
