Nvidia earnings in focus as DeepSeek threatens AI chip demand

When Nvidia reports its earnings on Wednesday, investor focus will be squarely on the company’s pricey artificial intelligence chips amid some stirring question marks about the long-term sustainability of large outlays on AI.

Nvidia has emerged as the biggest winner in a two-year investment boom driven by AI. The most successful technology companies have spent billions of dollars to purchase Nvidia’s advanced chips to run their new AI programs. But there are fresh questions about whether heavy spending is warranted.

This idea has been irreversibly flipped by the recent arrival of China’s DeepSeek and its low-price AI models. DeepSeek has claimed its models have made performance success similar to those of the best Western AI systems but for a fraction of the price.

These claims could turn the current AI hardware landscape, reducing the reliance on Nvidia’s advanced chips. Now, some investors wonder whether Nvidia’s hold on A.I.-infrastructure companies is quite as unassailable as it once seemed. Without this supply, companies may have to spend money on more costly options, which could dampen Nvidia’s soaring revenue growth.

Nvidia investors are highly concerned about the effect of DeepSeek on their investment

Nvidia’s next earnings report will be scrutinized for its financial results and signs of shifting demand trends for AI chips. DeepSeek’s meteoric rise in January sent Nvidia’s market value crashing down by a jaw-dropping $593 billion, which became the largest one-day loss on record for an American company. Nonetheless, Nvidia shares were among the very top in 2023 and 2024.

DeepSeek is a new genre of artificial intelligence that decodes data at never-before-seen speeds, which concerns Nvidia’s investors. 

Chart displays rise in Nvidia’s market capitalization

If Nvidia could demonstrate its ability to “beat and raise,” it would likely benefit stocks. According to LSEG data, Nvidia was projected to report fourth-quarter revenue of $38.05 billion, reflecting a 72% increase, marking its slowest growth in seven quarters. 

The company was also expected to forecast a 60% revenue surge for the first quarter ending in April. This would represent a departure from Nvidia’s previous trend, as the company had experienced five consecutive quarters of triple-digit revenue growth through the period ending in October.

There are competitive issues, but demand for Nvidia’s AI chips appears robust. Nvidia’s biggest customers, such as Microsoft and Meta, have signaled that they would recommit to heavy investments in data centers, indicating that demand for its technology won’t slow down anytime soon.

At the same time, the sales of Nvidia’s powerful Blackwell chips likely helped revenue in the fourth quarter, though those high-yield chips are squeezing profit margins due to the corporate expenses needed to scale up a new, complex chip. Analysts expect Nvidia’s adjusted gross margin to decline over three percentage points to 73.5% for the quarter.

Chart represents slow-down in Nvidia’s revenue growth

Nvidia is no longer just selling individual chips like the GB200 NVL72—thanks to the Blackwell series, it’s now selling complete AI computing systems. These systems combine GPUs, CPUs, and networking components to provide customers with a more robust, high-performance AI infrastructure.

Blackwell chips are made using an advanced packaging process, which is more complicated and costly. Taiwan Semiconductor Manufacturing Company (TSMC), the world’s largest contract manufacturing manufacturer’s primary supplier, has been racing to increase its capacity for this complex process of coalescing multiple chips to boost performance and efficiency. 

Demand is far greater than available capacity for advanced packaging, which has become a critical choke point in AI semiconductor supply chains, adding pressure to supply.

Community designs, including the SoC and DaVinci, etc., were initially marred by design issues and low chip yields, resulting in poor performance. These issues affected initial production runs, delayed rollout, and, at one point, led to concerns over the stability of the supply.

However, Nvidia has addressed those problems, fine-tuning the design and increasing production yields, resulting in a more reliable and economical delivery timeline. Despite these early setbacks, Nvidia is still optimistic about the Blackwell series to solidify the company’s position as a leader in AI hardware. It is also purpose-built to provide the strength and speed required to run the most demanding AI workloads and position the company to seize growing opportunities in the fast-evolving AI landscape.

Nvidia’s performance might fall short of the exceptional results 

Nvidia may miss the very good results. In November, Nvidia said that revenue from its fourth quarter Blackwell series of chips would likely be several billion dollars higher than previous guidance thanks to high demand from its new generation of AI chips. 

However, the rollout has been anything but simple, with challenges from manufacturing and supply chain constraints to making advanced technologies work together in end-to-end AI computing systems.

Gabbeli’s top Belton underlined the hurdles of bringing Blackwell to market, calling it an especially tricky line to launch. He added that investors began to expect Nvidia to outperform expectations. However, due to Blackwell’s production and logistical challenges, the company’s outperformance may not be as pronounced this time.

Advanced packaging constraints, early design difficulties, and a shift towards selling complete AI computing systems instead of discrete chips have also added to factors that could temper the size and shape of Nvidia’s traditional earnings beat. However, analysts and investors say Nvidia can overcome those obstacles and maintain its dominance in the AI hardware market.

It’ll be interesting to see how the company’s next earnings report fits into this story, how much is learnable about how well Blackwell and the subsequent generations of Nvidia architectures are being absorbed, and whether Nvidia can maintain its legacy of delivering rapid expansion to its shareholders as competitive pressures intensify and the markets for its solutions change.

Cryptopolitan Academy: Coming Soon - A New Way to Earn Passive Income with DeFi in 2025. Learn More