2032 AI investment wave will break through $500 billion! Google has the most advantage, Microsoft, Meta still limited by Nvidia

Bloomberg, a well-known news agency, released a report on the AI industry on 3/17, stating that global tech giants' investment in AI will exceed $525 billion in 2032, which is $154 billion more than this year's forecast.

This wave of AI investment is mainly dominated by mega-scale computing companies such as Microsoft, Amazon, and Meta. It is expected that the total investment in AI this year will reach as high as $371 billion, a 44% increase from 2024, demonstrating the high dependence of global enterprises on AI technology and the white-hot market competition.

AI investment model transformation, shifting from training to inference

In the past, technology companies mainly invested in AI funds for training large AI models, that is, collecting and processing massive data to develop new technologies. However, with the rise of reasoning models such as OpenAI and Chinese DeepSeek, there has been a noticeable shift in the direction of corporate fund allocation.

With the development of inference models, technology companies may shift more costs to the operational stage after AI goes online, rather than the initial training, causing drastic changes in the AI investment structure.

DeepSeek impacts the market, low-cost and efficient operation becomes the focus of industry attention

Chinese AI startup DeepSeek from China previously claimed to have developed competitive AI models at significantly lower costs than US AI companies. The news immediately sparked discussions and even prompted major US companies to reevaluate the rationality of their substantial AI investments.

The report pointed out that some technology companies have begun to shift to a 'more efficient' AI computing method to reduce operating costs. It is expected that the demand for future AI hardware may not grow as crazily as in the past few years, and enterprises will pay more attention to how to make AI operate more efficiently at lower cost.

2032 The proportion of inference technology in AI total investment is 50%, and the investment structure of AI has changed significantly

Currently, about 40% of the AI budgets of major tech companies are still used for model training, but this proportion may decrease to only 14% by 2032. On the other hand, investment in AI 'inference' technology is rapidly increasing, with the expectation that by 2032, inference technology will account for over 50% of total AI investment, becoming the fastest-growing market in the AI field.

Google has the upper hand, while Microsoft and Meta are restricted by Nvidia

In this AI transformation war, Google has the most competitive advantage due to its self-developed AI chips, which can handle AI training and inference simultaneously, with higher flexibility. However, Microsoft (Microsoft) and Meta mainly rely on Nvidia (Nvidia) chips, which may lead to a lack of sufficient autonomy when the AI computing paradigm shifts, affecting the layout of future AI strategies.

Efficiency is the top priority, which tech giants can win

The AI industry is undergoing a major change in the flow of capital and the application of technology, shifting from the past 'competitiveness in computing power and training' to an AI operation mode oriented towards 'efficiency.' With DeepSeek challenging traditional models and rationalizing the hardware requirements for AI, how technology giants adapt to this AI revolution will be the key to victory.

(DeepSeek cannot just spend 6 million euros! AI encryption czar analyzes the source of gold behind: China uses Singapore to obtain NVIDIA chips)

This article predicts that the 2032 AI investment wave will exceed 500 billion US dollars! Google has the most advantage, while Microsoft and Meta are still limited by Nvidia. Originally appeared on ChainNews ABMedia.