There is no single, comparable “AI revenue” total in the NVIDIA and Microsoft filings discussed here. Both companies report broader business categories and describe AI as a growth driver, but neither provides a harmonized, standalone AI-revenue figure in those disclosures. Hardware sales are real revenue; the challenge is that large equipment shipments and cloud services have different scopes, economics and places in the AI supply chain.
What counts as AI revenue?
Use a disclosure ladder rather than treating every AI-related figure as equivalent:
As an Amazon Associate I earn from qualifying purchases.
- Defined AI-specific revenue: Call a figure AI revenue only when a company identifies a defined AI line or gives a transparent calculation. The NVIDIA and Microsoft filings cited here do not provide a harmonized standalone total.
- Reported segment or end-market: Preserve the issuer’s label. NVIDIA’s Compute & Networking segment and Data Center end-market are reported categories, not interchangeable measures and not automatically “AI revenue.”
- AI-attributed growth: A company may say AI helped drive a category’s growth. That explains a driver; it does not mean every dollar in the category came from AI.
- Third-party estimate: Treat an outside estimate separately, identifying the estimator, method, date and scope. The filings below do not validate an external estimate.
This distinction matters because a broad segment can include products and activity beyond AI, while an end-market label and a segment label may use different classifications. A reader cannot calculate the exact AI share of those categories from the figures alone.
What NVIDIA and Microsoft actually reported
The figures below come from different fiscal years, so this is a comparison of disclosure types, not a synchronized ranking. NVIDIA’s fiscal 2026 ended January 25, 2026; Microsoft’s fiscal 2025 ended June 30, 2025.
#1 Best Overall
- NVIDIA Volta GV100 Architecture — 4,608 CUDA Cores, 640 1st-Gen Tensor Cores delivering 14 TFLOPS FP32 and 112 TFLOPS deep learning performance for AI training, inference, HPC, and scientific computing workloads
- 32GB HBM2 ECC Memory — 900 GB/s Bandwidth — High-bandwidth memory on a 4096-bit bus with ECC error correction provides the memory capacity and throughput required for the largest AI models, simulations, and datasets
- PCIe 3.0 x16 Interface — 250W TDP — Standard PCIe Gen3 connectivity with passive cooling designed for enterprise rack server deployment in HPE ProLiant, Dell PowerEdge, and Supermicro platforms with adequate chassis airflow
- NVLink — Scale to 96GB Unified Memory — Connect two V100 GPUs via NVLink at 300 GB/s bi-directional bandwidth to scale GPU memory from 32GB to 96GB for larger AI training and HPC workloads
- Multi-Precision Computing — Supports FP64 (7 TFLOPS), FP32 (14 TFLOPS), FP16 (112 TFLOPS) and INT8 precision modes for flexible deployment across training, inference, and scientific simulation workloads
| Company and period | Reported measure | Reported result | What it does—and does not—show |
|---|---|---|---|
| NVIDIA, fiscal 2026 | Total revenue | $215.938 billion, up 65% year over year | Company-wide revenue; not an AI-only total. |
| NVIDIA, fiscal 2026 | Compute & Networking segment | $193.479 billion, up 67% | A reported segment whose increase NVIDIA linked partly to platform shifts including accelerated computing and AI; not a standalone AI figure. |
| NVIDIA, fiscal 2026 | Data Center end-market | $193.737 billion, compared with $115.186 billion in fiscal 2025 | A separately reported end-market measure. Do not substitute it for, or add it to, the similarly sized Compute & Networking segment. |
| Microsoft, fiscal 2025 | Intelligent Cloud revenue | $106.265 billion, up 21% | A reported cloud segment, not a standalone AI-revenue total. |
| Microsoft, fiscal 2025 | Azure and other cloud services growth | 34% | Growth in a cloud-services category; not the growth rate of a defined AI-revenue line. |
NVIDIA said Data Center computing grew 59% and Data Center networking grew 142% in fiscal 2026. It attributed Data Center computing growth to demand for its Blackwell platform; those component growth rates do not redefine the entire Data Center end-market as AI revenue. NVIDIA’s fiscal 2026 Form 10-K and Microsoft’s 2025 Annual Report provide the underlying company disclosures.
Why hardware sales can make growth look unusually fast
A large shipment of equipment can add substantial revenue when customers expand data-center capacity. If those purchases ramp quickly, a hardware supplier’s reported growth may look especially steep compared with a business whose revenue is recognized through cloud services or software subscriptions over time. The rates are not directly comparable unless the underlying revenue types, periods and accounting treatment are understood.
Rank #2
- Delivers 25 TFLOPS of AI inference performance using four MemryX MX3 accelerators in a compact M.2 2280 M-key module.
- Connects through PCI Express Gen 3 and above for straightforward integration into embedded and edge systems with an M.2 M-key slot.
- Includes 42 MB of distributed on-chip weight storage to support low-latency, batch-1 pipelined inference.
- Designed for low-power deployment with a 1 W–12 W operating range and host-managed airflow requirements.
- Supports ONNX, TensorFlow, Keras, and TFLite models through the MemryX software stack on Linux and Windows hosts.
There is also a supply-chain issue: a hardware company can record a sale to a direct customer, while a cloud provider may later earn service revenue using infrastructure that includes such equipment. Those are different transactions at different points in the chain. Adding their reported revenues as if they represented distinct end-user AI spending can double-count economic activity; the filings do not provide a consolidated total that resolves that overlap.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Hardware growth is not inherently a distortion or an accounting error. It is a reason to ask what is being counted, how concentrated the sales are, and what costs accompany the growth.
Rank #3
- ✅Powered by 26 Tera-Operations Per Second (TOPS) Hailo-8 AI Processor. 2.5W typical power consumption
- ✅Scalable, enabling simultaneous processing of multi-streams & multi-models
- ✅Enabling real-time, low latency and high-efficiency AI inferencing on the edge devices
- ✅Supports TensorFlow, TensorFlow Lite, ONNX, Keras, Pytorch frameworks
- ✅Supports Linux and Windows. Supports the temperature range of -40°C to 85°C
Revenue growth is not the same as growth quality
Margins and costs
NVIDIA’s gross margin was 71.1% in fiscal 2026, down from 75.0% in fiscal 2025. The company cited the transition to Blackwell full-scale data-center solutions and a $4.5 billion H20 inventory and purchase-obligation charge. The revenue increase therefore should not be read as proof that margins remained unchanged.
Microsoft reported a 69% Microsoft Cloud gross margin percentage for fiscal 2025 and linked the decline to scaling AI infrastructure, partly offset by efficiency gains in Azure. Intelligent Cloud revenue, Azure growth and Microsoft Cloud gross margin each describe a different aspect of the business; none alone is an AI-revenue measure.
Rank #4
- 48GB AI graphics accelerator
Customers and attribution
NVIDIA said two direct customers accounted for 22% and 14% of its total revenue in fiscal 2026. It also cautioned that indirect-customer revenue is estimated and can differ from actual results. Direct-customer concentration and estimated end-customer attribution affect how confidently readers can connect a supplier’s sales to final AI demand.
Geography and periods
NVIDIA reports geography based on the headquarters of direct customers and cautions that end-customer location and shipping location can differ. A reported geographic split therefore should not be treated automatically as the location of final demand. Fiscal years also differ across issuers, so growth figures should be compared only after aligning the periods and understanding each company’s measure.
Best Value
- Professional AI & Creator Workstation: AMD Radeon AI PRO R9700 GPU with 32GB GDDR6 is engineered for AI development, professional content creation, and compute-intensive workloads.
- Massive 32GB Memory Capacity: 32GB of GDDR6 memory on a 256-bit bus provides ample bandwidth for large AI models, 8K video editing, and complex 3D rendering.
- Advanced RDNA 4 with AI Accelerators: 64 Compute Units with 3rd Gen Ray Tracing and dedicated 2nd Gen AI Accelerators for groundbreaking AI performance and visual computing.
- Professional Blower Cooling: Efficient single blower design exhausts heat directly out of the chassis, ideal for multi-GPU workstation and server configurations.
- Enterprise-Grade Thermal Solution: Vapor chamber heatsink with industrial Honeywell PTM7950 thermal interface material ensures reliable cooling under sustained professional loads.
How to assess a claim about AI-driven growth
- Find the exact reported line. Is it a defined AI product or service, a broad segment, an end-market, or a management explanation of a growth driver?
- Check the scope and period. Note the fiscal year, year-over-year basis, and whether the measure is a segment, end-market or service category.
- Identify the revenue type. Distinguish equipment shipments from subscription or consumption-based services; do not assume their timing or recurrence is identical.
- Read growth alongside economics. Check gross margin, cost pressures, inventory charges and infrastructure investment where disclosed.
- Consider who is counted. Look for direct-customer concentration, reseller or channel sales, estimated end users and possible overlap between suppliers and service providers.
- Keep labels intact. If the filing reports a broad category and says AI contributed to growth, describe it as AI-attributed growth—not as total AI revenue.
These checks help answer “How much of a company’s growth comes from AI?” without claiming a precision the filings do not provide. In the cited disclosures, the exact AI share of the broader reported categories is not established.
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.
