Global emissions from electricity consumption tied to artificial intelligence (AI) chip manufacturing surged last year, driven by heavy reliance on fossil fuels in Asia’s leading production technology hubs, according to Greenpeace East Asia.
Advertisement
Electricity consumption for AI chip production more than tripled to about 984 gigawatt hours (GWh), resulting in a fourfold increase to 453,600 metric tons of carbon dioxide emissions, Greenpeace said in a report on Thursday. Major hubs like Taiwan, South Korea and Japan face big climate and pollution risks if they keep relying on dirty fuels for power, it added.
“While fabless hardware companies like Nvidia and AMD are reaping billions from the AI boom, they are neglecting the climate impact of their supply chains in East Asia,” co-author Katrin Wu said in a statement. “AI chipmaking is being leveraged to justify new fossil fuel capacity in Taiwan and South Korea – demand that could, and should, be met by renewable energy sources.”
Demand for AI chips boomed as ChatGPT started an evolution in AI services and sparked a rise of Chinese rivals, banking on processors specially designed to perform big tasks like machine learning. These chips are significantly faster and more efficient than traditional computer processing units.
The global semiconductor market could grow 15 per cent this year, driven by the strong demand for advanced chips to handle AI workloads, according to market researcher IDC. The AI market could reach US$780 billion to US$990 billion in 2027, according to consultancy Bain & Co.
Advertisement