Nvidia reported $81.6 billion in revenue for fiscal Q1 2027, up 85% from a year earlier and the highest quarterly figure in its history. Data center sales reached $75.2 billion of that total. This post covers Nvidia’s 2026 revenue, AI chip demand, market share, and GPU compute performance across the Hopper, Blackwell, and Rubin generations.
Nvidia Statistics 2026 – TL;DR
- Nvidia posted record revenue of $81.6 billion in Q1 fiscal 2027, an 85% jump year over year.
- Data center revenue hit $75.2 billion, up 92% from a year earlier and 21% sequentially.
- Nvidia holds roughly 86% of the AI accelerator market, ahead of AMD and Intel.
- The B200 delivers about 20 PFLOPS of FP4 compute, a precision format the H100 does not support.
- AI inference token generation rose tenfold in a single year, according to CEO Jensen Huang.
The Nvidia statistics for 2026 point to one driver above all others: AI chip demand from data centers. Full-year fiscal 2026 revenue reached $215.9 billion, with data center sales of $193.7 billion making up about 90% of the company. Blackwell now accounts for most of that revenue, and the Rubin platform enters production in the second half of 2026. The numbers below break down revenue, demand by customer type, market share, and the GPU compute metrics that separate each chip generation. For wider context on where the sector is heading, Quantumrun tracks the trajectory of artificial intelligence across industries.
How Much Revenue Does Nvidia Make in 2026?
Nvidia’s revenue climbed every quarter through fiscal 2026 and accelerated into fiscal 2027. Total revenue grew from $44.1 billion in Q1 fiscal 2026 to $81.6 billion in Q1 fiscal 2027.
| Quarter | Total Revenue | Data Center | YoY Growth |
|---|---|---|---|
| Q1 FY26 | $44.1B | $39.1B | 69% |
| Q2 FY26 | $46.7B | $41.1B | 56% |
| Q3 FY26 | $57.0B | $51.2B | 62% |
| Q4 FY26 | $68.1B | $62.3B | 73% |
| Q1 FY27 | $81.6B | $75.2B | 85% |
Source: NVIDIA quarterly results (Forms 8-K)
Non-GAAP gross margin held at 75% in Q1 fiscal 2027, up from 71.3% a year earlier. Nvidia guided to about $91 billion in revenue for the following quarter.
Nvidia Data Center Revenue and AI Chip Demand
Data center is the core of Nvidia’s business in 2026. The segment generated $193.7 billion for full-year fiscal 2026, up 68%, and accounts for roughly 90% of company revenue. Demand splits between hyperscalers and a widening base of AI clouds, enterprises, and sovereign buyers.
| Segment (Q1 FY27) | Revenue | YoY Growth |
|---|---|---|
| Hyperscale | $37.9B | 115% |
| AI Clouds, Industrial & Enterprise | $37.4B | 74% |
| Edge Computing | $6.4B | 29% |
Source: NVIDIA CFO Commentary, Q1 Fiscal 2027
Hyperscalers hold steady at about half of data center revenue. Data center networking revenue reached $14.8 billion in Q1 fiscal 2027, up 199% year over year, on demand for NVLink, Spectrum-X Ethernet, and InfiniBand.
AI chip demand shows up in deployment commitments too. Nvidia cited roughly 5 million GPUs across announced AI factory projects, a 10 gigawatt buildout with OpenAI, and an initial 1 gigawatt with Anthropic. Inference token generation rose tenfold in a year, a signal that demand stretches well beyond model training. These commitments line up with the technology milestones expected by 2027.
What Is Nvidia’s AI Chip Market Share in 2026?
Nvidia holds about 86% of the AI accelerator market measured by revenue, ahead of AMD and Intel. The company derives 91% of its revenue from data centers. Analysts at SemiAnalysis and Mercury Research project Nvidia’s share could ease toward 75% by the end of 2026 as custom hyperscaler ASICs and AMD’s MI-series scale.
| Supplier | Approx. AI Accelerator Share |
|---|---|
| Nvidia | ~86% |
| AMD | 5-8% |
| Custom ASICs and others | ~8% |
Source: Mercury Research / SemiAnalysis via The Motley Fool
The total market is expanding faster than any single rival can capture, so Nvidia’s revenue keeps rising even as its percentage share is forecast to slip. The CUDA software stack and priority TSMC CoWoS packaging allocation remain central to that lead. The longer arc of this shift is covered in Quantumrun’s long-range technology forecasts.
Nvidia GPU Compute Performance Metrics
Each Nvidia generation widens the gap on memory and low-precision compute. The H100 set the 2022 baseline. Blackwell doubled transistor count and added native FP4. Rubin, in production for the second half of 2026, moves to HBM4 memory.
| GPU | Architecture | Transistors | Memory | Bandwidth | FP4 Compute |
|---|---|---|---|---|---|
| H100 | Hopper (2022) | 80B | 80GB HBM3 | 3.35 TB/s | Not supported |
| B200 | Blackwell (2024) | 208B | 192GB HBM3e | 8 TB/s | ~20 PFLOPS |
| B300 | Blackwell Ultra | 208B | 288GB HBM3e | 8 TB/s | ~20 PFLOPS |
| R100 | Rubin (2026) | 336B | HBM4 | ~22 TB/s | ~50 PFLOPS |
Source: Spheron, NVIDIA product specifications
The practical effect is throughput. B200-based systems ran GPT-3 175B pre-training about 2x faster than H100 systems, and the GB200 NVL72 rack generates up to 30x more inference tokens than an equivalent Hopper setup. FP4 also shrinks model memory: a 70-billion-parameter model needs about 140GB in FP16 but only 35GB in FP4.
Why FP4 Precision Matters
Lower precision lets one GPU serve more inference copies at once. A single 192GB B200 can run four FP4 copies of a 70B model where an H100 in FP16 runs none. That math is what drives revenue per GPU-hour for cloud providers and feeds the broader rollout of predictive AI systems.
Nvidia Statistics 2026: Capital Returns and Outlook
Nvidia returned about $20.0 billion to shareholders in Q1 fiscal 2027 through buybacks and dividends. The board added $80.0 billion to its repurchase authorization and raised the quarterly dividend from $0.01 to $0.25 per share.
China remains a swing factor. Nvidia shipped no data center Hopper products to China in Q1 fiscal 2027, against $4.6 billion a year earlier, and assumed no China data center revenue in its outlook. For how compute demand could evolve next, Quantumrun maps emerging tech trends shaping the decade ahead.
FAQs
How much revenue did Nvidia make in 2026?
Nvidia reported $215.9 billion in revenue for full-year fiscal 2026, up 65%. In the following quarter, Q1 fiscal 2027, revenue hit a record $81.6 billion, up 85% year over year.
What is Nvidia’s AI chip market share?
Nvidia holds about 86% of the AI accelerator market by revenue as of early 2026, ahead of AMD and Intel. Analysts expect that share to ease toward 75% by year-end as custom ASICs scale.
How much faster is Blackwell than the H100?
B200 systems run GPT-3 175B pre-training about 2x faster than H100 systems. For inference, the GB200 NVL72 rack produces up to 30x more tokens than an equivalent Hopper configuration.
How much of Nvidia’s revenue comes from data centers?
Data centers account for roughly 90% of Nvidia’s revenue. Full-year fiscal 2026 data center revenue was $193.7 billion, and Q1 fiscal 2027 data center sales reached $75.2 billion.
When does Nvidia’s Rubin GPU launch?
Rubin entered full production for availability in the second half of 2026. First cloud deployments include AWS, Google Cloud, Microsoft, and Oracle, plus CoreWeave, Lambda, Nebius, and Nscale.
Citations
https://www.sec.gov/Archives/edgar/data/0001045810/000104581026000051/q1fy27pr.htm
https://nvidianews.nvidia.com/news/nvidia-announces-financial-results-for-fourth-quarter-and-fiscal-2026
https://www.fool.com/research/data-center-revenue-ai-companies/
https://www.spheron.network/blog/nvidia-rubin-vs-blackwell-vs-hopper/
