In January, a China-based startup called DeepSeek launched an artificial intelligence (AI) chatbot that rapidly became the most downloaded free mobile application in the US. The remarkable aspect of this development was the claim that the chatbot’s underlying large language model had been trained for only $6 million and outperformed certain leading US models in specific tests. Consequently, investors began selling their shares in Nvidia (NASDAQ: NVDA) under the belief that American companies were overspending on AI infrastructure. In a single day, Nvidia’s shares lost 17% of their value, resulting in the company’s market value decreasing by around $600 billion. This marked the largest one-day loss in the history of the US stock market. Following the turbulence caused by DeepSeek, Nvidia’s shares are still trading approximately 9% below and falling short of the all-time high by roughly 13%. However, Nvidia shareholders found some relief from positive news coming from Amazon (NASDAQ: AMZN) and Alphabet (NASDAQ: GOOGL). Both Amazon and Google’s parent company announced last week their fourth-quarter financial results, stating that due to high demand for AI infrastructure, they would significantly increase capital spending by 2025. Amazon CEO Andy Jassy foresees that the cost of training AI models will continue to decrease over time, making it easier for more companies to integrate “productive AI” techniques. Amazon CFO Brian Olsavsky mentioned that due to the demand for AI infrastructure, capital expenditures, which amounted to $83 billion in 2024, could exceed $100 billion by 2025. Alphabet CEO Sundar Pichai stated that with the drop in training costs, a larger portion of capital spending would shift to the inference infrastructure. Alphabet CFO Ana Ashkenazi predicted that due to investments in data centers, servers, and network infrastructure, capital expenditures, amounting to $52 billion in 2024, are expected to reach $75 billion by 2025. Amazon and Alphabet’s investment plans support the idea that demand for Nvidia GPUs may increase as training costs decrease. Andy Jassy drew a comparison to the increase in demand despite lower cloud computing prices, saying: “Companies reduce their costs per unit of infrastructure, lowering their expenses. Yet, they turn to projects they previously couldn’t implement due to high costs. So, even if cost per unit decreases, total technology spending increases. I think the same will happen with AI.” Expectations for Nvidia’s Substantial Gains According to Wall Street analysts, Nvidia’s adjusted earnings are projected to grow by an average of 52% annually until the end of the 2026 fiscal year in January 2026. This forecast makes the company’s current approximately 50-level adjusted P/E (Price/Earnings) ratio relatively attractive. Of course, these expectations are quite high, posing the risk of steep declines in Nvidia’s shares if the company fails to meet this performance. In conclusion, Nvidia shareholders shaken by concerns related to DeepSeek may find some relief. Most analysts believe that more efficient training methods will increase demand for Nvidia GPUs. The statements from major cloud infrastructure providers also support this view.