NVIDIA'S Stocks Tumble on Disappointing Forecasts, Blackwell Chips Snag

Nvidia Corp. recently reported its latest financial results, but the tech giant’s performance didn’t quite meet the lofty expectations of investors. The company delivered an underwhelming forecast and disclosed challenges in producing its highly anticipated Blackwell chips, leaving some investors disappointed.

‍

Nvidia’s quarterly report, arguably the most eagerly awaited in the tech industry, managed to meet or exceed analysts’ estimates across almost all metrics. However, Nvidia investors have grown accustomed to exceptional results, and this latest report fell short of those high standards.

‍

A key issue is the production difficulties surrounding Nvidia’s next big revenue driver, the Blackwell processor lineup. These chips, which represent the next generation of Nvidia’s leading artificial intelligence processors, have proven more difficult to manufacture than initially expected.

‍

As a result, Nvidia’s shares dropped by 2% in early trading in New York. This decline followed a year of incredible growth for the stock, which had more than doubled by Wednesday’s close, after surging 239% in 2023. The company’s impressive run had set the bar high, leading to what Bloomberg Intelligence analysts Kunjan Sobhani and Oscar Hernandez Tejada described as “unsustainable expectations.”

‍

For the third quarter, Nvidia projected revenue of about $32.5 billion. Although this figure slightly exceeded the average analyst prediction of $31.9 billion, it fell short of some of the more optimistic estimates, which had gone as high as $37.9 billion. This forecast has raised concerns that the AI-driven excitement propelling Nvidia’s rise might be cooling down. Nvidia has been the primary beneficiary of a massive push to upgrade data centers for AI software, and its sales projections have become a key indicator of this ongoing trend.

‍

Ahead of the earnings announcement, there were already rumors that Nvidia was encountering issues with its new Blackwell chips. The company confirmed these concerns, admitting that it faced production challenges and was making adjustments to improve its manufacturing yield—the percentage of usable chips produced. Despite these setbacks, Nvidia expressed confidence that the Blackwell chips would generate “several billion dollars” in revenue during the fourth quarter.

‍

Nvidia's CEO, Jensen Huang, reassured investors that supply issues would be resolved as production ramps up. He emphasized that once manufacturing gains momentum, Nvidia would have plenty of supply to meet demand.

‍

Nvidia has consistently delivered blockbuster results in recent quarters, often exceeding even the most optimistic Wall Street estimates. However, the magnitude of these surprises has been gradually decreasing. A significant portion of Nvidia’s revenue growth has been driven by a small group of customers, particularly large data-center operators like Google (part of Alphabet Inc.) and Meta Platforms Inc. These companies have been investing billions into AI infrastructure, fueling Nvidia’s rapid rise.

‍

While companies like Meta have increased their capital expenditure budgets this earnings season, there is growing concern that the current pace of infrastructure development might be excessive. Some fear this could lead to a bubble, though Huang remains optimistic, insisting that we are only at the dawn of a new era for technology and the economy.

‍

Nvidia has been the standout performer in the S&P 500 Index this year, far outpacing gains by other semiconductor companies. With a market value exceeding $3 trillion, Nvidia is worth nearly as much as the next 10 largest chipmakers combined.

‍

Initially known for its video-game graphics cards, Nvidia has since become a leader in AI accelerators—chips adapted from its graphics processors that are essential for training artificial intelligence models. These models become better at recognizing and responding to real-world inputs through a process called training, where AI is bombarded with information. Nvidia’s chips are also crucial for the subsequent phase, known as inference, where the trained AI models are deployed in real-world applications, such as powering services like OpenAI’s ChatGPT.

‍

In the most recent quarter, Nvidia’s results surpassed Wall Street’s expectations once again. The company reported revenue of $30 billion for the fiscal second quarter, more than double the previous year’s figure. Excluding certain items, Nvidia’s profit came in at 68 cents per share, exceeding analysts’ predictions of 64 cents per share.

‍

Nvidia’s success in the AI space has given it a significant head start over competitors, though rivals like Advanced Micro Devices Inc. (AMD) and Intel Corp. are trying to catch up. However, the combined revenue from AMD and Intel’s AI-related products is still only about 5% of Nvidia’s total.

‍

Nvidia’s data-center division, now its largest revenue source, generated $26.3 billion last quarter, while its gaming chip segment brought in $2.9 billion. Both figures exceeded analysts’ expectations.

‍

The upcoming Blackwell chips are expected to drive another wave of growth for Nvidia as they are rolled out in the coming months. Despite the delays, analysts remain confident in the strong demand for Nvidia’s current generation of products, which could help the company weather any production hiccups without a significant financial impact.

‍

During a conference call following the earnings report, analysts pressed for more details on the expected revenue from the new Blackwell chips and their timeline for availability. However, Huang and Nvidia’s Chief Financial Officer, Colette Kress, remained tight-lipped, sticking to their forecast of billions in revenue for the fourth quarter without providing further specifics.

‍

As the call progressed, Nvidia’s stock extended its losses, reflecting the market’s frustration with the lack of detailed answers. Nonetheless, Huang remained focused on the future, predicting that the ongoing AI revolution would require a trillion dollars’ worth of new equipment to replace outdated data-center infrastructure. This massive replacement cycle, he argued, is just beginning, underscoring the long-term potential of AI in transforming industries and economies.

‍