Nvidia’s revenue forecast on Wednesday rattled Wall Street, sending its stock down 2.5% in after-hours trading. The company projected its slowest revenue growth in seven quarters, raising questions about whether the artificial intelligence (AI) boom that catapulted Nvidia to prominence might be losing steam. However, analysts, investors, and Nvidia’s leadership firmly reject that notion, pointing instead to a relentless demand for Nvidia’s cutting-edge chips and persistent supply chain constraints as the key culprits behind the tempered forecast.
The surge in demand for Nvidia’s graphics processing units (GPUs), indispensable for training and running AI models, shows no signs of slowing down. As companies across industries scramble to adopt generative AI technologies, Nvidia’s GPUs, particularly its flagship H100 and A100 series, have become the gold standard for AI infrastructure.
“We’re seeing unprecedented demand for our chips,” said Nvidia CEO Jensen Huang during a conference call with investors. “Companies are eager to deploy generative AI systems, and we are working tirelessly to meet their needs.”
Despite this optimism, Nvidia faces a significant hurdle: it can only sell chips as fast as its manufacturing partner, Taiwan Semiconductor Manufacturing Co (TSMC), can produce them. TSMC’s advanced manufacturing capabilities are vital to Nvidia’s operations, but even they are struggling to keep pace with the demand.
One of the key challenges Nvidia faces is the advanced packaging required for its latest chip, codenamed Blackwell. Unlike traditional chips, Blackwell consists of multiple smaller chips integrated into a single unit through a complex process known as advanced packaging. While this design improves performance, it also complicates manufacturing.
“Blackwell adds more advanced packaging from TSMC than prior chips, which adds a wrinkle,” said Ben Bajarin, CEO and principal analyst at Creative Strategies. Packaging has become a significant bottleneck, not just for Nvidia but across the semiconductor industry.
TSMC is racing to expand its packaging capacity, but the process involves specialized facilities and techniques, making rapid scaling difficult. Analysts expect these constraints to persist well into 2025.
Adding to Nvidia’s challenges, a design flaw discovered in the Blackwell chip earlier this year necessitated a “mask change” — a costly and time-consuming adjustment. This issue lowered chip yields, meaning fewer chips emerged from the production line fully functional. Although Nvidia has resolved the flaw, it has set back production timelines.
“There’s the risk that the bottlenecks worsen rather than improve, and that could damage revenue projections,” said Michael Schulman, chief investment officer at Running Point Capital.
The mask change highlights the complexity of manufacturing chips like Blackwell, which involve hundreds of intricate steps. These steps include photolithography, where ultraviolet light passes through masks to etch circuits onto silicon wafers. Any disruption in this process can have cascading effects on production schedules.
Nvidia’s revenue forecast reflects these supply chain constraints. While the company remains confident in its long-term growth, the near-term outlook shows signs of strain. Nvidia executives expect gross margins to drop to the low-70% range as the company ramps up Blackwell production, a decline from its previous margins in the mid-70% range.
“In the short term, the production ramp-up will pressure gross margins,” Huang explained. “But this is a temporary phase as we iron out production kinks.”
Despite the lowered growth projections, Nvidia maintains that demand remains robust. The company has already shipped approximately 13,000 samples of its Blackwell chip and expects to surpass initial revenue estimates for the current quarter.
The global semiconductor industry has been grappling with supply chain disruptions for years, and Nvidia’s situation underscores the fragility of these systems. The shortage of advanced packaging capacity is particularly problematic because it affects Nvidia’s ability to capitalize on booming AI demand.
Hendi Susanto, a portfolio manager at Gabelli Funds, emphasized that the company’s future hinges on its ability to scale production. “The key focus is supply — how much supply Nvidia can produce,” he said. “Demand is absolutely and exceptionally strong for the foreseeable future.”
TSMC’s efforts to expand its advanced packaging facilities will play a critical role in alleviating these bottlenecks. However, the timeline for these improvements may not align with the soaring demand Nvidia faces.
The AI boom has been a transformative force for Nvidia, driving its market capitalization to over $1 trillion earlier this year. Companies ranging from tech giants to startups are racing to deploy generative AI systems, and Nvidia’s GPUs are at the heart of this revolution.
While other players in the semiconductor industry, such as AMD and Intel, are attempting to compete, Nvidia’s dominance in the AI chip market remains unchallenged. Its CUDA software platform, coupled with its state-of-the-art hardware, provides a comprehensive ecosystem for AI developers.
Yet, the question remains: how long can Nvidia sustain its leadership position if supply chain constraints persist? Analysts believe that the company’s ability to navigate these challenges will be crucial to maintaining its edge.
Nvidia’s disappointing revenue forecast has sparked mixed reactions from investors. While some are concerned about the immediate financial implications of supply chain issues, others remain optimistic about the company’s long-term prospects.
“Nvidia is in a unique position,” said Susanto. “Even with these challenges, it remains the go-to supplier for AI infrastructure. The demand is there; it’s just a matter of meeting it.”
Investors will be closely watching how Nvidia manages its production ramp-up and whether TSMC can deliver the necessary capacity expansions in time. The stakes are high, not just for Nvidia but for the broader AI ecosystem, which relies heavily on its technology.
