Hyperscalers Increase Spending on AI Infrastructure, Boosting NVIDIA's Prospects
Recently, major tech players known as hyperscalers, including Amazon, Microsoft, Alphabet, and Meta, released their financial results for the third quarter. Along with these results, they also shared insights into their capital expenditure (capex) plans for the upcoming years, particularly in relation to the growing field of artificial intelligence (AI).
The news was exciting, as analysts at Morgan Stanley highlighted that the total capex for these hyperscalers in 2025 could surpass $300 billion. This figure represents a significant increase of 25% compared to the planned spending for 2024 and is almost twice the amount spent in 2023.
Due to this surge in spending, Morgan Stanley analysts raised their capex forecasts for 2025 by 9%. Amazon and Microsoft are expected to take the lead in this expenditure, planning to invest $96.4 billion and $89.9 billion, respectively. According to analyst Brian Nowak, these investments are pivotal as the companies continue to leverage opportunities in General AI and large language models (LLMs).
Nowak further noted the importance of sharing information regarding how these companies are adopting and engaging with new AI technologies, as well as the financial returns from those investments. As the demand for sophisticated AI capabilities rises, tracking these disclosures will become essential.
The primary beneficiary of this massive rise in spending is anticipated to be NVIDIA. The company specializes in high-end AI chips, which will be crucial to support the expanded infrastructure that hyperscalers are developing.
According to BofA Securities analyst Vivek Arya, NVIDIA is projected to secure over 75% of the AI accelerator market share through 2027. Other competitors, such as Advanced Micro Devices Inc. and Broadcom Inc., are expected to capture only a small portion of this market.
Arya expressed optimism about the growth of the AI accelerator market, estimating it will reach $280 billion by the end of 2027 and could potentially increase to over $400 billion in the long run. This is a dramatic rise from the approximately $45 billion estimated for 2023.
In summary, the increased capital expenditures planned by hyperscalers for AI-related technologies indicate a strong future for NVIDIA and highlight the expanding role of AI across various sectors.
AI, NVIDIA, Hyperscalers