Nvidia Extends AI Growth on Data Center Demand, Vera Chips

Key Moments

Investor Reaction and Outlook

May 20 (Reuters) – Nvidia CEO Jensen Huang spoke on Wednesday. He aimed to reassure investors about the company’s growth outlook. He said Nvidia can scale further through a broader customer base and new AI products.

However, Nvidia shares still fell 1.6% in after-hours trading. This drop reflected investor caution about rising competition. Even so, the company delivered strong guidance for Q2. It also announced an $80 billion buyback plan.

Nvidia expects Q2 revenue of $91 billion, plus or minus 2%. In contrast, analysts at LSEG had forecast $86.84 billion.

In addition, Nvidia raised its quarterly dividend to 25 cents per share. Previously, it was just 1 cent per share.

AI Demand and Market Position

Nvidia plays a central role in the AI industry. Its chips power most global data centers. As a result, its performance is closely watched by investors.

“Nvidia delivered another beat, but that is now expected,” said eMarketer analyst Jacob Bourne. He added that the key question is durability through 2027 and 2028. He also noted rising competition from Google, Amazon, AMD, and Intel.

Meanwhile, AI infrastructure spending continues to grow rapidly. For example, Alphabet, Amazon, and Microsoft may spend over $700 billion this year. That compares with about $400 billion in 2025.

Hyperscalers and New Data Center Customers

On the analyst call, Huang said Nvidia expects faster growth than its biggest hyperscale customers. In addition, he pointed to rising demand from AI-focused cloud providers.

These smaller cloud firms are growing faster than traditional hyperscalers. At the same time, their sales are now comparable in size.

“We should grow faster than hyperscale capex,” Huang said.

Competitive Threats From Custom and Rival Chips

Nvidia still dominates AI hardware. However, many customers are now building custom chips. This shift creates new competitive pressure.

At the same time, Intel and AMD are targeting AI inference workloads. Both companies see a large growth opportunity in this segment.

To respond, Nvidia launched a new CPU and AI platform in March. It also integrated technology from startup Groq to improve inference performance.

Vera CPUs Open Additional Revenue Stream

Huang said Nvidia’s “Vera” CPUs open a new $200 billion market. He also said the company expects $20 billion in revenue from Vera this fiscal year.

This forecast is separate from Nvidia’s earlier $1 trillion projection for Blackwell and Rubin chips.

“I expect Vera to be the second-largest contributor,” Huang said. He also noted strong customer interest in the platform.

However, supply constraints remain a concern. Huang said demand may exceed supply through the Vera Rubin product cycle.

Supply Chain, Capex and Cloud Agreements

Nvidia is increasing spending to avoid supply bottlenecks. It is responding to a global shortage of memory chips.

As a result, its supply commitments rose to $119 billion in the first quarter. That is up from $95.2 billion in the previous quarter.

Quarterly Results Snapshot

Nvidia reported strong Q1 results. Revenue came in at $81.62 billion. Analysts had expected $78.86 billion, according to LSEG.

Data center revenue also beat expectations at $75.2 billion. The estimate stood at $72.8 billion.

On an adjusted basis, Nvidia earned $1.87 per share. Analysts expected $1.76 per share.

Cloud Commitments to Support R&D

Nvidia also reported $30 billion in cloud agreements. This is up from $27 billion last quarter.

These agreements support its research and development efforts. In some cases, they also act as financial backstops for cloud partners.

According to Seaport analyst Jay Goldberg, these deals may help manage excess capacity. However, they also reflect Nvidia’s deep integration with cloud infrastructure providers.