Nvidia's $1 trillion chip sales forecast faces supply and competition risks

Nvidia's $1 trillion cumulative chip sales target through 2027 remains on track after data center revenue grew 92 percent to $75 billion in the fiscal first quarter, though supply constraints and customer-built silicon pose risks.

"We saw $500 billion of very high confidence demand and purchase orders for Blackwell and Rubin through 2026," CEO Jensen Huang said at the company's GTC conference in March. "I'm here to tell you that right now where I stand ... I see through 2027 at least $1 trillion."

In fiscal 2026, which mostly aligns with calendar 2025, Nvidia generated $216 billion in total revenue, with nearly $194 billion from data centers. Wall Street's consensus estimate forecasts total revenue of $394 billion this year and $561 billion next year, about $1.17 trillion in cumulative revenue since last year. If data centers remain roughly 90 percent of total revenue, cumulative data center sales from Blackwell and the new Rubin chips would exceed $1 trillion by the end of next year.

CFO Colette Kress reaffirmed the target on the fiscal first-quarter earnings call, saying the company is "working vigorously on our supply chain ecosystem to address the incredible demand." That comment also flags the two biggest risks: whether Nvidia can ship enough chips to meet demand, and whether its largest customers shift more workloads to their own custom AI silicon.

Supply Chain Bottlenecks Test Execution

AI demand has created a bottleneck stretching from data center construction to memory and other components needed to build advanced chips. Nvidia relies on TSMC for wafer fabrication and on a complex packaging supply chain for its HBM-equipped accelerators. The Blackwell platform uses TSMC's advanced process nodes and CoWoS packaging, which has been a persistent constraint across the AI chip industry. Demand for Nvidia's hardware looks enormous, but the open question is whether the company can ship enough to meet it. Kress's emphasis on supply chain work suggests management sees the constraint as the primary gating factor.

The scale of the challenge is significant. Nvidia's data center revenue alone reached $75 billion in a single quarter, and the company is expected to nearly double that run rate as Rubin ramps. Each generation of accelerators requires more advanced packaging, more HBM memory, and more power delivery infrastructure — all constrained across the industry. TSMC has been expanding CoWoS capacity, but lead times for advanced packaging remain extended. Memory suppliers including SK Hynix and Micron have also been ramping HBM production to keep pace with demand from Nvidia and other AI chip designers.

Customers Building In-House Chips

A second risk comes from Nvidia's own customers. Amazon and Alphabet's Google are designing custom AI chips for their cloud platforms, and top cloud companies have previously made up about half of Nvidia's data center revenue. Amazon's Trainium and Google's TPU lines have been gaining traction for inference workloads, where cost efficiency matters more than raw training performance. If those customers shift more workloads to in-house silicon, Nvidia's growth could slow.

That's why Nvidia has been striking partnerships with neocloud and sovereign customers, including IREN and leading Japanese manufacturers, to reduce dependence on big tech buyers. These partnerships diversify the customer base and create demand outside the hyperscaler ecosystem. The strategy appears to be working — Nvidia's order book includes $500 billion in high-confidence demand through 2026, according to Huang. The company has also been expanding its networking and software offerings, which add recurring revenue streams beyond one-time chip sales.

Investor Takeaway

Nvidia shares trade at roughly $200.75, up 2.93 percent, with a market capitalization near $4.9 trillion. The $1 trillion cumulative sales target is still in play, but it isn't guaranteed. If the data center market slows or supply constraints limit Nvidia's ability to fulfill orders, growth and the stock could face pressure. Investors tracking Nvidia's quarterly data center revenue and supply chain commentary will get the next read when the company reports fiscal second-quarter results. The stock's valuation already reflects substantial growth expectations, so any shortfall against the $1 trillion trajectory could trigger a meaningful repricing. For the broader AI semiconductor complex, Nvidia's execution on this target serves as a barometer for the entire AI infrastructure buildout.

This article is for informational purposes only and does not constitute investment advice.