Apple supplier Foxconn, the world’s largest contract electronics maker, has adopted a far more bullish outlook for this year. On Thursday, it expected a significant rise in revenue driven by booming demand for artificial intelligence servers. The upbeat tone came after a forecast-beating 33% jump in net profit for the fourth quarter. It also contrasted sharply with remarks by Foxconn Chairman Young Liu in November that the world’s largest contract electronics maker had a tough time predicting a recovery in consumer product demand this year amid ongoing economic uncertainties and geopolitical tensions.
Liu’s upbeat stance on 2024 comes even as the company expects revenue for intelligent consumer electronics to decline in the first quarter from a year earlier and for computing, cloud, networking, and component products to remain flat this year. That could pressure the company’s share price by 0.5% in early trading.
The company, which assembles the iPhone and other top-selling electronic devices for Western markets and China, also plans to expand into electric vehicles and build an ‘AI factory’ globally to create new jobs and technologies. It collaborates with Nvidia Corp to build a manufacturing system using intelligent robots and software. It aims to replicate its success with smartphones and tablet computers with EVs.
In addition to expanding its global footprint, the firm focuses on AI and other high-growth markets, including virtual reality and self-driving cars. Liu said the EV business was already “reaching and being approached by many automakers.”
It is building an “AI Factory” in Taiwan, which will create a new type of manufacturing anchored on artificial intelligence. The ‘AI Factory’ will involve building and running intelligent robots and systems that use machine learning to create parts for product assembly. It will work with both traditional and start-up automakers, he added.
He added that the ‘AI Factory’ will help reduce costs and improve quality. It will allow companies to focus more on their core businesses, enable them to make more accurate predictions about sales, inventory, and other business variables, increase productivity and efficiency, and improve product quality.
Researchers have been raising general alarms about AI’s hefty energy requirements. Still, a study published this week in Joule Journal is one of the first to quantify the potential impact. It estimates that if AI servers ran at total capacity, they would consume 85.4 terawatt-hours per year, more than many small countries use yearly.