Nvidia AI Chips Face Rising Competition
Nvidia AI chips face pressure from hyperscaler custom chips and AMD as H200 China shipments start, forcing traders to weigh guidance and backlog.
KEY TAKEAWAYS
- Hyperscalers' custom inference chips threaten Nvidia's revenue concentration; they account for 40-50% of revenue.
- China H200 approval allows 40,000-80,000 unit shipments implying $1.28-2.56B potential FY2027 revenue.
- FY2026 guidance of $213B and current-quarter $65B reflect scale and a near $500B backlog.
HIGH POTENTIAL TRADES SENT DIRECTLY TO YOUR INBOX
Add your email to receive our free daily newsletter. No spam, unsubscribe anytime.
On Jan. 26, 2026, Nvidia began shipping its H200 AI chips to China as the company faces growing competition from hyperscalers’ custom inference chips and price-competitive rivals. This shift could alter Nvidia’s revenue mix and investor outlook.
Market Share and Competitive Pressures
Nvidia controls about 80–90% of the high-end AI chip market and roughly 85–90% overall, supported by its CUDA software stack and NVLink interconnect, which create a strong ecosystem moat. This dominance has helped Nvidia lead sales of top-tier AI accelerators.
Major hyperscalers—Microsoft, Meta, Amazon, and Google—account for 40–50% of Nvidia’s revenue. They are increasingly deploying their own custom chips for inference tasks, which represent an estimated 80% of long-term AI compute demand. This trend reduces external GPU demand and could erode Nvidia’s revenue concentration, even as it remains the preferred supplier for many workloads.
Outside hyperscalers, AMD held about a 7% share of the AI accelerator market in the third quarter of 2025, up 0.8 percentage point. Its MI300X chip offers a roughly 20–30% lower-cost alternative on comparable workloads, creating pricing pressure in some segments.
Revenue Guidance, Backlog, and China Shipments
Nvidia reported third-quarter 2025 revenue of $57 billion, up 62% year over year. Data-center sales reached $51.2 billion, while networking revenue rose 162% to $8.2 billion, reflecting strong demand from cloud and enterprise customers.
The company projects fiscal-year 2026 revenue of $213 billion, a 63% increase, with consensus earnings per share of $4.69, up 57%. Current-quarter revenue guidance stands at $65 billion. Nvidia’s order backlog approached $500 billion for 2025–2026, including a data-center backlog estimated at $275 billion based on prior shipments. Gross margins remain in the 73–75% range, supported by high-margin products and software.
China approved Nvidia’s H200 chip in January 2026, enabling initial shipments of 40,000–80,000 units priced around $32,000 each. This could generate low-billion-dollar revenue in fiscal 2027. Under a U.S. agreement, 25% of Nvidia’s Chinese revenue from these sales will be payable to the U.S. government. Production capacity is expanding to meet orders for the second quarter of 2026. The lifting of prior U.S. restrictions in April 2025 and this approval have shifted the geography and economics of Nvidia’s China sales.
Broader demand assumptions include an estimated $500 billion in hyperscaler AI capital expenditures for 2026, which could sustain large-scale orders even as these companies develop more custom inference hardware. Meanwhile, ongoing U.S. Department of Justice and European Union antitrust probes into Nvidia’s bundling practices may affect how the company packages chips with networking and software.
