Can Nvidia Beat EPS Forecasts While Losing Ground in China?

On Wednesday, May 28th, Nvidia (NASDAQ:) is scheduled to report its Q1 earnings ending April 2025. To beat investor expectations, the dominant AI chip designer would have to beat the analyst forecast of $0.8 earnings per share (EPS), per Zacks Investment Research.

In the year-ago quarter, Nvidia reported an of $0.58, which was 625% higher than the EPS reported in Q1 2023. But now that the company has an enormous market weight of $3.2 trillion, the expected yearly EPS uptick is a moderate 38%.

Year-to-date, NVDA stock performance is in the negative yield territory, at -5%, having its Relative Strength Index (RSI) fall under 70 last Tuesday. Over a month, however, NVDA shares gained 26.5% value, as it became clear that President Trump’s “Liberation Day” regarding the imposition of global tariffs has dubious leverage.

Ahead and beyond Wednesday’s earnings report, what can investors expect from NVDA stock exposure?

Nvidia’s Diminishing Prospects in China

When the Chinese startup launched its eponymous DeepSeek R1 model, it surprised the West. Chinese developers managed to harness Nvidia’s H800 chips, as nerfed versions of H100 GPUs, to deliver higher efficiency and lower computing costs with clever parallelism and compression techniques.

Nvidia’s H800s were the company’s answer to the federal government’s first wave of chip export controls implemented in October 2022. Since then, it has been a bipartisan consensus to stifle China’s access to cutting-edge chips with even greater export controls.

In other words, the AI race mirrors geopolitics between the US-aligned West and China. In early May at the Milken Institute, US Treasury Secretary Scott Bessent noted that the US has to “win in AI and Quantum”. Otherwise, “if we don’t win, everything else doesn’t matter.”

For fiscal year 2025, China generated $17.11 billion in revenue for Nvidia. If we take into account the smuggling of AI chips via Singapore, Nvidia’s sales in that region would account for 33.26% of the company’s total revenue for that period.

Expectedly, Nvidia CEO Jensen Huang is not pleased by escalating chip controls to target China. Most recently at Computex Q&A, Huang noted that Nvidia had to exercise “writing off multiple billions of dollars” just on H20 series restriction alone. In total, Huang regrettably informed the audience that the Biden admin lowered Nvidia’s market share in China from 95% to 50%.

Huang also observed that such restrictions would only accelerate China’s chip design and manufacturing ecosystem. Indeed, we have concluded that Huawei, in concert with SMIC, is likely to nullify Western advantage by 2030 or earlier.

Offsetting Factors for Nvidia’s Bottom Line

Although aggressive chip export controls are concerning for Nvidia’s bottom line, as a global purveyor of AI tech through TSMC, it is also likely that AI demand will offset it. From the very beginning of the AI hype, we’ve kept reminding readers that Nvidia’s compute power is yet to see a demand uptick once text-to-video outputs become more than a novelty.

In just two years, Google’s Veo 3 video generator crossed that major milestone recently, as it is now exceedingly difficult to discern the difference between AI content and video recordings. Video generation, combined with audio generation, requires drastically more compute power.

Not only is each frame equivalent to text-to-image output, but all frames have to have temporal consistency. Additionally, video generation mimics physics modeling to make interactions and object motion indistinguishable from real footage.

To offer greater control in AI apps of the future, alongside robotics, we will likely see a hybrid approach with physics-informed loss functions. This fits perfectly into Nvidia’s full-stack offering of GPUs and AI accelerators that simulate physics, such as Newton. Case in point, Flexcompute announced in March that integrating Nvidia’s Blackwell platform made it possible to achieve 100x faster physics simulations.

On a large scale, Nvidia will also be a key player in supplying tech for smart cities in the form of public-private partnerships (PPPs), as we’ve recently seen with Saudi Arabia’s AI investment spree. Specifically, to enable decision-making in AI-powered governance systems, Nvidia’s Omniverse Cloud can run “physical AI solutions with digital twins”.

Exerting its hegemony, the USG has already made inroads with the Stargate project, announced shortly after President Trump’s inauguration. Most recently, on May 22nd, UAE announced the deployment of Stargate UAE cluster utilizing Nvidia’s GB300 platform, in joint collaboration with OpenAI and Oracle (NYSE:).

For the USG allowing Nvidia tech in the UAE, the Emirates committed up to $1.4 trillion AI infrastructure investments in the US. In short, Nvidia should not only be viewed as the dominant semiconductor designer company, but as an integral part of US efforts to secure its technological and geopolitical hegemony.

We are likely to see many more PPPs that benefit Nvidia’s bottom line in the future. After all, as we’ve maintained all along, AI deployment is primarily about erecting new types of governance systems.

NVDA Price Targets

Presently priced at $131.29 per share, the average NVDA price target sits at $161.82, according to WSJ’s forecasting data. The bottom estimate is $100, while the ceiling is all the way up to $235.92 per share.

Only one analyst recommends selling, while the overwhelming majority, 52 analysts, view NVDA stock as a buy opportunity at this price point.

***

Neither the author, Tim Fries, nor this website, The Tokenist, provide financial advice. Please consult our website policy prior to making financial decisions.

This article was originally published on The Tokenist. Check out The Tokenist’s free newsletter, Five Minute Finance, for weekly analysis of the biggest trends in finance and technology.