Direct Answer
60 SEC filers are classified under the Standard Industrial Classification code that describes the gpu & ai accelerator semiconductors industry, and their latest annual reports show $542.8 billion of combined revenue and a median operating margin of -2.3%. Investors should still evaluate its revenue model, industry-specific KPIs, valuation framework and the risks that can change the group's economics.
Industry Overview
GPU and AI accelerator makers design the chips that train and run artificial intelligence models, plus the software and systems around them. Revenue is concentrated in hyperscalers, and supply, software ecosystem and product cadence decide share. The first numbers to compare across companies here are data center revenue and growth, gross margin and backlog and supply.
Key Industry Metrics
- Data center revenue and growth: Sales of accelerators and related systems and their growth. It shows how much results depend on hyperscaler spending, and high growth can reflect both volume and pricing.
- Gross margin: Revenue left after wafers, memory and packaging. Leading products earn high prices, while competition from custom silicon and supply costs can reduce it over time.
- Backlog and supply: Orders on hand and the ability to deliver them. Supply of advanced packaging and memory limits shipments, so backlog shows demand and supply sets the pace.
- Customer concentration: The share of sales from a few cloud buyers. Their capital plans drive orders, so a high share leaves results sensitive to a single large customer's decisions.
- Software ecosystem adoption: How many developers and customers build on the company's programming tools. A strong ecosystem raises switching costs, since software moves slowly to rival chips.
Primary Macro Drivers
- AI model training and inference demand: More models trained and run in production require accelerators in large numbers, driving data center revenue, backlog and supply allocation.
- Hyperscaler capital spending: Cloud providers' budgets decide how many accelerators are bought, appearing in data center growth and backlog.
- Packaging and memory supply: Chips can only ship when advanced packaging and memory are available, so these inputs limit revenue and affect gross margin.
- Export rules: Rules limiting sales of advanced chips to certain countries can remove demand or force redesigned products, shifting regional revenue.
Major Risks
- Customer concentration: A few cloud and AI buyers account for most revenue, so a change in their plans shows directly; watch top-customer share and backlog.
- Capex digestion: Buyers can pause orders while absorbing earlier purchases, causing a sharp revenue slowdown; watch backlog, customers' capital guidance and inventory.
- Custom silicon competition: Large buyers building their own chips can move volume away from merchant suppliers; watch customer concentration, pricing and disclosed share trends.
- Export restrictions: New limits can block shipments to certain customers or regions, removing demand; watch regional revenue mix and license requirement disclosures.
Industry Characteristics
| Characteristic | Profile |
|---|---|
| Cyclicality | Cyclical |
| Capital Intensity | Low |
| Regulatory Intensity | Medium |
| Industry Maturity | Growth |
| Geographic Importance | Global |
GPU & AI Accelerator Semiconductors Filers in the SEC Record
60 companies with annual reports in the dataset are classified by the SEC in the Standard Industrial Classification code 3674. Their latest fiscal years end between 2025 and 2026, and together they reported $542.8 billion of revenue. Median revenue growth over the prior year was 20.3%. Median operating margin was -2.3% and median net margin -1.7%. Median total liabilities were 34% of total assets, and 27 of 60 reported a profit in their latest year.
| Company | Revenue | Growth | Operating margin | Net margin | Year |
|---|---|---|---|---|---|
| Nvidia Corp | $215.9 billion | 65.5% | 60.4% | 55.6% | FY 2026 |
| Broadcom Inc. | $63.9 billion | 23.9% | 39.9% | 36.2% | FY 2025 |
| Intel Corp | $52.9 billion | -0.5% | -4.2% | -0.5% | FY 2025 |
| Micron Technology Inc | $37.4 billion | 48.9% | 26.1% | 22.8% | FY 2025 |
| Advanced Micro Devices Inc | $34.6 billion | 34.3% | 10.7% | 12.5% | FY 2025 |
| Applied Materials Inc | $28.4 billion | 4.4% | 29.2% | 24.7% | FY 2025 |
| Texas Instruments Inc | $17.7 billion | 13.0% | 34.1% | 28.3% | FY 2025 |
| Nxp Semiconductors N.V. | $12.3 billion | -2.7% | 24.8% | 16.5% | FY 2025 |
The 8 largest of 60 by revenue. Source: each company's own annual report as filed with the SEC (XBRL data), figures in U.S. dollars. A company's SIC code reflects its main business as the SEC records it, so it can differ from how an index provider or analyst classifies the company, and a diversified group appears under only one code.
Where to Go Next
The GPU & AI Accelerator Semiconductors guide describes how the industry earns money, what to measure and what can go wrong. To see how it compares, read the neighbouring industries in the same sector, then apply the same checks to a specific company.
More in Information Technology
- Information Technology sector overview
- High-Performance Computing as a Service
- Gaming Infrastructure Software
- Human Capital Management Software
- Foundation Model Developers
- Identity & Access Management
- Financial Software
- Industrial Electronics
- Fabless Semiconductors
Put it to work
Frequently Asked Questions
What are the main KPIs for the GPU & AI Accelerator Semiconductors industry?
Key performance indicators for the GPU & AI Accelerator Semiconductors industry include: Data center revenue and growth, Gross margin, Backlog and supply, Customer concentration, Software ecosystem adoption. Which matter most depends on the company's model, so compare them across several companies rather than relying on one.
How cyclical is the GPU & AI Accelerator Semiconductors industry?
The GPU & AI Accelerator Semiconductors industry is cyclical, meaning its revenues and earnings tend to expand during economic expansions and contract during downturns. Investors should monitor leading economic indicators and interest rate trends when evaluating timing.
What are the biggest risks in the GPU & AI Accelerator Semiconductors industry?
The primary risks for GPU & AI Accelerator Semiconductors companies include: Customer concentration, Capex digestion, Custom silicon competition, Export restrictions. Investors should assess each company's exposure to these factors and its track record of navigating them.