Nvidia’s latest quarterly revenue has more than doubled compared to previous results, and the company anticipates that its third-quarter revenue will exceed Wall Street predictions, which indicates a sustained demand for its AI chips.
The firm is projecting revenue growth of around 70% for fiscal 2028 and also announced that it will be expanding its partnership with Amazon Web Services, the cloud computing arm of Amazon.
According to Nvidia’s finance chief, Colette Kress, they plan to deploy an additional 2 million Nvidia GPUs within Amazon’s worldwide infrastructure by 2027 and 2028.
In after-hours trading, Nvidia’s shares climbed over 4%, reversing earlier trends, as investors expressed optimism regarding Nvidia’s future growth strategy.
As of the close on Wednesday, Nvidia shares have risen by 12.4% this year, while competitors AMD and Intel have seen even greater increases, with their shares more than doubling.
Nvidia is often viewed as a bellwether for the AI sector, given that its chips are integral to many major data centers and advanced AI models globally.
This announcement comes on the heels of expectations from big players like Microsoft and Meta, two of Nvidia’s top clients, who anticipate that the tech industry will invest over $730 billion in AI infrastructure this year, a significant increase from last year’s $400 billion.
However, a growing portion of funds designated for tech spending is now being directed towards in-house chip development, which aims to lessen dependence on Nvidia’s expensive and limited supply processors.
Add to that, Nvidia’s role in the intricate financing landscape of the AI boom is under more scrutiny after the company agreed to back certain deals with six major financial institutions, aiming to acquire over $500 billion for AI infrastructure.
Analysts are particularly focused on Nvidia’s profit margins, anticipating that they might face strains due to the ramp-up in production for its Rubin chips coupled with rising memory prices affecting the overall silicon supply chain.
Nvidia estimates its adjusted gross margin for the third quarter will be around 74%, with a small range of variability, while analysts had predicted 74.77%.
For the upcoming third quarter, the company expects revenue of $108 billion, with a 2% margin of error, in contrast to the average analyst estimate of $104.19 billion, based on LSEG data.
An important note is that Nvidia has not accounted for any data center chip sales in China within its forecasts, given the uncertainty surrounding their operations in the region. Back in May, Washington cleared about 10 Chinese companies, like Alibaba and Tencent, to purchase Nvidia’s powerful H200 AI chips, though actual deliveries have faced delays.
Nvidia has started promoting its new Vera CPU to Chinese customers since June, indicating availability by August, while discussions continue about permitting limited H200 purchases for top AI firms in China.
Recently, a U.S. Commerce Department official mentioned that although shipments had commenced, they were still relatively minimal.
As AI increasingly automates various tasks and addresses queries, Nvidia’s graphics processors are contending with heightened competition from central processors and specialized chips that are better adapted for inference tasks.
This competitive landscape is driving many tech companies to invest in their own semiconductor design.
For instance, Meta is expected to start the production of its custom “Iris” AI chip in September, part of a broader initiative to create four generations of custom silicon aimed at reducing computing expenses.
An additional report indicated that Alphabet has placed orders for over 3 million chips from Intel for 2028 and that Nvidia is considering leveraging Intel’s manufacturing technology to develop a processor combining four GPUs into one unit.
Intel and AMD, among other rivals, are also looking to capitalize on the inference market, while several Chinese companies, including Baidu, are already developing their own chips for similar applications.
