Text size
Nvidia has seen soaring demand for its data center products that enable generative AI projects.
Justin Sullivan/Getty Images
Nvidia
is approaching the trillion-dollar-value club after the chip maker’s latest revenue forecast sent the stock soaring.
The company cited surging demand for its products that enable the latest wave of artificial intelligence applications.
Investors are now looking for other opportunities to benefit from the AI trend, which has already sparked Nvidia stock’s triple-digit gains this year. Ironically, the best ideas for the potential next big winners might have been mentioned by the CEO of Nvidia himself.
On Wednesday, the semiconductor company reported April quarter earnings results moderately above expectations. But the big news was the guidance. For the current quarter, Nvidia gave a revenue forecast range of $11 billion at the midpoint, well above the analyst consensus of $7.2 billion.
The company’s management said the upside came from soaring demand for its data center products that enable generative AI projects. OpenAI’s release of ChatGPT late last year sparked interest in this form of AI. The technology ingests text, images, and videos brute-force to create content. And Nvidia’s chips are the market leader in making that technology possible.
Nvidia said it saw broad-based AI demand from cloud computing providers, large consumer internet companies, start-ups, and enterprises. It suggests many enterprises and upstarts are experimenting with AI, concerned they risk being caught lagging behind their competitors.
The company’s stock surged 27% to $386.71 on Thursday, the day after the report, achieving a market value of roughly $950 billion.
At this level, however, the potential for further gains is more difficult and entails higher risk. The market is embedding a significantly higher growth trajectory for Nvidia in the coming years. The chip maker trades 55 times this fiscal year’s projected earnings, or 24 times sales estimates of $39 billion—an extremely elevated valuation versus other prominent industry players.
In comparison,
Amazon.com
(AMZN) is valued at $1.2 trillion with an estimated revenue of $560 billion this year, while Intel (INTC) is worth $114 billion with a full-year revenue consensus of $51 billion.
The issue is some of the new AI ventures won’t prove sustainable. Not every app or website needs a chatbot, and not every generative AI start-up will find a business model. And because generative AI requires expensive computing power and server costs, only the projects that can add value and lead to paying customers will survive over time.
That isn’t to say there won’t be success stories. On the earnings call Wednesday, Nvidia CEO Jensen Huang may have pointed to two good prospects. He repeatedly expressed excitement over Nvidia’s partnerships with
ServiceNow
(NOW) and
Adobe
(ADBE) as examples of companies extending generative AI capabilities to their corporate customers in a smart way.
Both companies make sense as potential generative AI winners, with their respective platforms and customer bases.
ServiceNow
can use Nvidia’s technology to develop custom large-language models to extend productive new functionality across the enterprise. For example, the models can automate repeated human resources requests and customer-service tasks using internal databases. Unlike general-use chatbots that can be asked anything, commonly asked questions by employees are limited and have a verifiable correct answer or process. That means ServiceNow’s enterprise-focused AI models can more likely be fine-tuned to work effectively.
ServiceNow trades at about 12 times this year’s sales estimate and
Adobe
trades at 9 times sales, both lower than Nvidia’s valuation.
For Adobe, the software company is adding generative AI tools called Firefly to content-creation software that allows users to create and alter images based on text prompts. It enables faster iteration of new design ideas and could also lower contract artwork costs. Corporate customers might also decide to use Adobe because its AI was trained on images that can be used for commercial work, compared with start-ups that often train their services on copyrighted photos found across the internet.
Huang’s comments about ServiceNow and Adobe might turn out to be a gift. Both companies represent clear examples where generative AI can enable groundbreaking features that customers willing to pay for. They may be the best way to play the AI boom from here.
Write to Tae Kim at tae.kim@barrons.com
