This is not investment advice. The author has no position in any of the stocks mentioned. Wccftech.com has a disclosure and ethics policy.
A fall in spending by big tech firms on AI products could have serious consequences for the semiconductor industry, warns Doug Lefever, the CEO of Japanese semiconductor test equipment provider Advantech. Lefever shared his thoughts in an interview with the Financial Times, adding that the potential slowdown could be mitigated if consumer AI smartphones gain popularity in the market. AI spending could materialize in cycles, and once the current cycle ends, the industry could experience a shock, believes the executive.
The surge in AI spending in 2024 revolved around the capacity to train and test the models through mega data centers. These data centers have reshaped the paradigm for scale and energy consumption and forced big tech firms to consider alternative energy sources, such as nuclear power, for their computing facilities.
As a result, most of the spending in the AI industry focuses on data centers and has originated from big technology firms eager to build their computing infrastructure. However, the CEO of the world's largest chip testing equipment provider, Advantech, Doug Lefever, believes that big tech's AI spending could be part of a cycle that could end in a "vicious" bottom before picking up again.
In an interview with the Financial Times, Lefever shared that consumers' adoption of AI through their smartphones could help buffer the supply chain against this cyclical downtrend. He added that "may not last long and then it may go right back up," but refrained from comparing current AI spending trends to a bubble since the term "implies that it's [the spending] going to go away." While consumer AI devices are unlikely to use NVIDIA's products, their models and computations should be facilitated through data centers and associated infrastructure that uses the GPUs.
While the initial wave of AI smartphone spending has been slow, everyone is "waiting for the killer app with the AI handsets," believes the executive. If this application materializes, then the resulting wave of AI hardware demand could help buffer against any cyclical downtrends in spending by big tech on AI data centers.
Even though an AI spending slowdown hasn't crossed many investors' minds in 2024, some quarters have been worried about the impact of competition on chip designer NVIDIA's sales. NVIDIA's Blackwell GPUs, which are the industry leaders in performance, have been in short supply and have commanded high price tags. As a result, big tech has diverted some of its spending away from NVIDIA to rivals like Broadcom and Marvel and towards designing and developing in-house AI processors.
NVIDIA, whose shares are up by 859% since the start of 2023, is no stranger to turmoil in the GPU market. Its shares had dropped by 50% between late 2018 and early 2019 after a sluggish Bitcoin market meant that miners started unloading their used GPUs into the market. This impacted the demand for NVIDIA's GPUs which resulted in the firm losing more than $23 billion in market capitalization.