Nvidia Faces Decelerating Growth as Market Competition Forecasts Shift

Market Sentiment and the Question of a Bubble

The conversation surrounding Nvidia’s future has shifted from simple growth projections to concerns about market saturation and the viability of the broader AI trade. During a recent visit to CNBC headquarters, the discussion centered on whether the AI sector is currently in a bubble. For many investors, including those managing family portfolios, the decision to hold or sell Nvidia has become a pressing concern as the stock’s performance in 2026 has slowed compared to previous years.

Nvidia has undeniably rewritten the standards for growth at scale. Since the AI story accelerated in 2022, the company’s data center business scaled roughly 13 times, moving from $27 billion in total company revenue in fiscal year 2023 to approximately $216 billion in fiscal year 2026. However, analysts suggest that repeating this 13x growth from a $200 billion base is statistically improbable. As one source noted, the era of triple-digit annual growth is likely transitioning toward a more durable, albeit decelerating, pace of 15% to 20%.

Meta Platforms and the Shift Toward Custom AI Chips

A significant factor influencing investor perception in 2026 is the move by major hyperscalers to reduce their dependence on Nvidia’s expensive graphics processing units (GPUs). Recent reports indicate that Meta Platforms plans to begin manufacturing an in-house AI chip this September, aiming to bolster the AI features powering Facebook and Instagram. The company intends to increase its overall AI compute power to 14 gigawatts by 2027.

This shift toward custom, application-specific integrated circuits (ASICs) is designed to handle specific inference workloads more efficiently than general-purpose GPUs. Market research firm TrendForce anticipates a 45% jump in sales of these custom processors this year. Despite this, Nvidia remains deeply embedded in the ecosystem. In February, Meta expanded its agreement with Nvidia for the large-scale deployment of Nvidia CPUs and millions of Nvidia Blackwell and Rubin GPUs, as well as the integration of Nvidia Spectrum-X Ethernet switches for Meta’s Facebook Open Switching System platform.

Competitive Pressure from AMD and Market Share Dynamics

While Nvidia has controlled an estimated 80% to 90% of the AI chip market, credible competition is finally emerging. AMD has begun shipping its first rack-scale AI system, known as Helios, to major customers like Microsoft. This system is viewed as the first true competitor in years to Nvidia’s Blackwell and Rubin systems.

Company	Market Position / Strategy
Nvidia	Dominant share (approx. 74% in inference); high switching costs via CUDA.
AMD	Growing threat with rack-scale AI systems and committed demand from Meta and OpenAI.
Meta	Developing four generations of in-house chips to reduce GPU dependency.

Evaluating the Investment Thesis for 2026

Despite the challenges, some analysts maintain a bullish outlook on Nvidia’s earnings growth. The stock currently trades at 23 times forward earnings, which some argue is an attractive entry point given the company’s role as a core component of global AI infrastructure. If the company’s earnings per share reach $16.06 in the next three years, some projections suggest the share price could reach $345.

Ultimately, the decision to hold or sell depends on an investor’s tolerance for the transition from explosive, moonshot growth to a more mature, competitive market phase. As noted in the discussion at CNBC, the answer is rarely a simple binary choice. Investors are encouraged to look beyond the “hold or sell” question and instead evaluate whether the company’s long-term dominance in data centers remains intact as hyperscalers diversify their hardware stacks.