Over the past few years, investing in leading companies in the field of artificial intelligence has been an effective strategy for achieving excess returns. Nvidia (NVDA), as a key hardware provider, has once become a market favorite due to the significant advantages of its graphics processing units. However, with the evolution of technology and changes in the competitive landscape, doubts have arisen in the market about whether it can maintain its dominant position in 2026, especially when alternatives such as Alphabet’s custom tensor processing units emerge. If its market share is eroded, its stock price may come under pressure; on the other hand, if it can consolidate its position by leveraging the continuously expanding AI expenditures, it is likely to reach new highs.
Current concerns mainly stem from the shift in AI computing focus from model training to inference. Training complex models requires powerful computing capabilities, which is the traditional strength of Nvidia’s GPUs. However, the requirements for computing units during the inference phase are different, with relatively lower technological barriers, potentially creating opportunities for custom accelerators from companies like Broadcom. Recent news suggests that Meta Platforms, an important customer of Nvidia, is purchasing TPUs jointly developed by Alphabet and Broadcom. Some investors interpret this as a signal of substitution risk, but the actual situation may be more complex. Given that Nvidia’s GPUs are still in short supply, Meta’s procurement behavior is more likely a diversification supplement to the supply chain rather than a comprehensive replacement. Therefore, concerns about its market position being shaken may be overly exaggerated.
From a valuation perspective, based on the 2026 earnings expectations, Nvidia’s price-to-earnings ratio is within its recent fluctuation range, indicating that the market still holds high growth expectations for it. Analysts generally predict that its revenue growth rate for the next fiscal year could reach 48%, and the achievement of this goal largely depends on its continued competitiveness in the AI computing market. Although competition in the inference phase is intensifying, Nvidia’s comprehensive advantages in its software and hardware ecosystem remain significant, and it is fully capable of maintaining an important share in the rapidly growing market.
A more long-term perspective comes from the growth of the overall industry scale. It is estimated that by 2030, global annual capital expenditure in data centers could reach the trillions of dollars. Against this grand backdrop, even if its market share fluctuates slightly, Nvidia is still likely to benefit from the dividends brought by industry expansion, thanks to its technological accumulation and customer base. Overall, the current market doubts may actually present opportunities for strategic positioning. If Nvidia proves its adaptability and competitiveness again in 2026, its performance is likely to exceed the expectations of most investors, making it a continued leader and winner in the wave of artificial intelligence.
