Insights on Green Computing: Challenges, Trends, and Solutions from Ant Group and Academia

Ant Group recently hosted its first "Green Computing" competition, drawing attention from academia and industry. In a podcast interview, Ant Group's green computing lead He Zhengyu and Nanjing University researcher Gu Rong discuss why green computing has become a priority, the massive energy consumption of data centers, and the need for more efficient use of compute resources.

They explain that the rapid growth of cloud computing and big data has led to soaring electricity demand, with a large portion of power not directly contributing to data processing. Companies like Google, Microsoft, Huawei, and domestic firms are pursuing hardware and software solutions, such as low‑carbon data centers, energy‑aware hardware design, and software‑level optimizations.

Gu emphasizes that software improvements—especially cloud‑native resource schedulers, elastic AI frameworks, and efficient data‑center management—can significantly raise utilization rates from the typical 8‑15% to 35‑70%, reducing waste and emissions. He cites Google's internal studies showing 50‑60% average utilization and Ant Group's own progress.

The conversation also covers challenges such as low CPU utilization (6‑12% average), the need for better resource allocation, and the importance of cloud‑native, serverless architectures to dynamically match supply and demand. They highlight competition tracks like cloud‑native scheduling, remote‑sensing image recognition, and time‑series traffic prediction, noting innovative ideas such as using Tetris‑style algorithms for resource packing.

Looking ahead, they stress the role of education, talent cultivation, and open collaboration between industry and academia to drive green computing forward, while also pointing to emerging trends in AI model efficiency, specialized hardware, and sustainable system design.