Nvidia’s quarterly earnings have been must-watch events ever since its stock began a dramatic ascent in 2023, following OpenAI’s launch of ChatGPT and the start of the generative AI boom. Now that it’s the world’s most valuable public company, Nvidia’s results face intense scrutiny as investors seek reassurance that AI-driven capital spending remains justified.
Today’s after-the-bell announcement will come amid weeks of tech stock selloffs. But Nvidia stock edged higher this morning, as analysts predicted data center revenue, adjusted earnings per share and gross profit margin would rise.
Certainly, the company’s bona fides have not changed amid the recent stock roller coaster ride: Nvidia controls the vast majority of the market for GPUs, the chips used to train and run large AI models like ChatGPT and Anthropic’s Claude. Its CUDA software platform, which lets developers write code optimized specifically for Nvidia hardware, has become the industry’s de facto standard, reinforcing that dominance. Meanwhile, the company’s rapid cadence of new chip generations—from Hopper to Blackwell, with Rubin on deck—signals no intention of slowing down.
Still, investors will be watching closely for signs of strain. One key question is whether competitors are beginning to gain real traction. For example, days after committing to deploy millions of Nvidia GPUs, Meta this week announced a multibillion-dollar deal to buy chips from AMD. The agreement also gives Meta the option to take up to a 10% stake in AMD, echoing a similar investment AMD secured from OpenAI last October.
Cloud giants are also working to reduce their dependence on Nvidia—even as they remain among its largest customers. Amazon has begun deploying thousands of its own AI chips across a sprawling network of data centers in Indiana, where they are being used by Anthropic. Google, meanwhile, has struck a series of deals with Anthropic and is reportedly supplying chips for several of the startup’s new data centers in New York, Texas, and elsewhere.
A growing field of startups focused on chips built specifically for inference—the process of generating outputs from trained AI models—also poses a challenge. Nvidia has moved to hedge against that threat, entering a high-profile, non-exclusive licensing deal with one of the most well-known of these startups, Groq, that included bringing CEO Jonathan Ross and other staffers to Nvidia.
But other startup chipmaking rivals continue to push forward. Cerebras Systems has reportedly filed confidentially for a U.S. IPO once again. SambaNova recently raised $350 million in a round that included Intel, and the Dutch-based Axelera just raised $250 million. And MatX—founded by former Google TPU engineers—has secured $500 million in Series B funding co-led by Jane Street and Situational Awareness, claiming its processors could outperform Nvidia’s GPUs by as much as tenfold for large language models and inference by 2027.
The rise of AI agents—including tools like Anthropic’s Claude Code and OpenAI’s Codex, as well as the viral OpenClaw—has further sharpened the focus on inference. Unlike chatbots, agents are designed to run continuously, which requires even more computing power. This is raising new questions about which chips will power this next phase of AI at such a massive scale.
As the leader of the pack by a mile, investors want to continue to see Nvidia’s year-over-year revenue increases. But these days, Nvidia is also a proxy for the entire AI boom—and the company’s results will be read as a verdict on whether the AI spending boom is still on track, or starting to crack.
