On Aug. 26, Nvidia CEO Jensen Huang Will Provide an Update on Nvidia's $1 Trillion GPU Opportunity. But There's a Smaller Part of the Business Investors Shouldn't Overlook
Nvidia reports Q2 fiscal 2027 earnings on Aug. 26, with focus on Jensen Huang's $1 trillion GPU opportunity update. Investors are also advised to monitor the emerging CPU business driven by agentic AI, which Huang projects a $20 billion revenue potential this year. Analysts highlight growing CPU market share and TAM expansion as key factors alongside data center GPU performance.
Investors are gearing up for Nvidia's (NVDA -0.02%) fiscal 2027 second-quarter earnings, which will be released after the market closes on Aug. 26.
Nvidia is at the center of the artificial intelligence (AI) trade, which is itself at the center of the broader market right now. So all investors will be at least somewhat interested in how the company performs and what CEO Jensen Huang has to say about overall demand.
Nvidia-specific investors will likely focus closely on what Huang has to say about his previous projection that the company will generate $1 trillion in sales from its Blackwell and Vera Rubin graphics processing units (GPUs) between 2025 and calendar year 2027.
While it's a massive number, there is another emerging part of the business that investors shouldn't overlook.
Nvidia CEO Jensen Huang. Image source: Nvidia.
Nvidia becomes more prevalent in agentic AI
While Nvidia doesn't break out its revenue by specific chip model, all GPU revenue falls within the company's data center division, which now makes up the bulk of Nvidia's revenue.
Nvidia's fiscal year begins in late January or sometimes early February, so it is fairly close to the calendar year. In fiscal 2026, most of which occurs in calendar year 2025, Nvidia's data center division generated nearly $194 billion in revenue.
According to Visible Alpha, Wall Street consensus estimates expect data center revenue to be nearly $368 billion in fiscal 2027 and nearly $531 billion in fiscal 2028.
Adding up all three years comes to roughly $1.09 trillion, so heading into the upcoming report, investors will be razor-focused on whether fiscal 2027 and fiscal 2028 data center revenue projections increase or decrease.
NASDAQ: NVDA
Key Data Points
While the focus will and should be on the GPU business, investors should not overlook Nvidia's central processing unit (CPU) business, which is fairly new. CPUs, which power older consumer electronics such as phones and laptops, have seen a major resurgence due to agentic artificial intelligence.
AI agents are used to perform autonomous tasks with minimal human interaction. While GPUs continue to power the reasoning, CPUs are now viewed as critical for actually performing the task, whether it's planning how to carry it out or accessing external files or data sources.
Last quarter, Nvidia launched its Vera CPU, specifically built for agentic AI. On the company's earnings call, Huang said this opens up a $200 billion total addressable market (TAM), and Nvidia expects about $20 billion in CPU revenue this year alone, instantly making it one of the largest, if not the largest, CPU players.
"With respect to CPU use, an agent is essentially what people call a harness," Huang told analysts on the company's most recent earnings call. "And so the harness runs on CPU... The world has a billion human users. I sense that the world is gonna have billions of agents. Not today. I mean, we are gonna grow into it."
CPUs will become a bigger part of the market
It's not just Huang who believes the CPU opportunity is massive.
Intel's CEO, Lip-Bu Tan, has also publicly said that the ratio of CPUs to GPUs used for AI inference has already moved from 1:8 to 1:4 and could approach parity or better in the future.
In a research note from June, Bank of America analysts expect the CPU TAM to grow fivefold from $35 billion in 2025 to $170 billion by 2030.
Pegging down a GPU TAM is more difficult, but Bloomberg projected earlier this year that it could rise to $486 billion by 2033, assuming a 14% compound annual growth rate (CAGR). That means by 2030, it would have been $328 billion, still 92% higher than the CPU TAM. But who knows how things will change, especially if Tan is right?
Either way, investors should look for any updated guidance and comments from Huang on the CPU opportunity, as it looks to be an important driver for the company and AI trade as a whole moving forward.
