As the industrialization of AI progresses further, demand for computing power continues to operate at a high level. According to Wande data, as of October 28, out of 53 listed companies in the computing power leasing sector, 25 had disclosed their 2026 three-quarter reports. Among them, 4 companies reduced their losses year on year in the first three quarters, 2 reversed their losses year on year, and 7 had a year-on-year increase in net profit. Wu Zewei, a special researcher at Sushang Bank, said that part of the reason for the significant differentiation in the performance of listed companies in the computing power leasing industry was that corporate revenue grew rapidly, but was hampered by one-time revenue disturbances, asset depreciation and amortization, and short-term profits were pressured; the other part was that the company performed well in the first three quarters due to the implementation of high-quality long orders and iterative business models. In the context of continuous differentiation in performance, simply locking in long orders and reducing operating costs can only relieve profit pressure, making it difficult to break through the profit ceiling of the hardware leasing business. To this end, leading enterprises have begun to break out of the single computing power rental model, transform into integrated computing power service providers, and enhance their competitive advantage by relying on software platforms and full-link services.

As the industrialization of AI progresses further, demand for computing power continues to operate at a high level. According to Wande data, as of October 28, out of 53 listed companies in the computing power leasing sector, 25 had disclosed their 2026 three-quarter reports. Among them, 4 companies reduced their losses year on year in the first three quarters, 2 reversed their losses year on year, and 7 had a year-on-year increase in net profit. Wu Zewei, a special researcher at Sushang Bank, said that part of the reason for the significant differentiation in the performance of listed companies in the computing power leasing industry was that corporate revenue grew rapidly, but was hampered by one-time revenue disturbances, asset depreciation and amortization, and short-term profits were pressured; the other part was that the company performed well in the first three quarters due to the implementation of high-quality long orders and iterative business models. In the context of continuous differentiation in performance, simply locking in long orders and reducing operating costs can only relieve profit pressure, making it difficult to break through the profit ceiling of the hardware leasing business. To this end, leading enterprises have begun to break out of the single computing power rental model, transform into integrated computing power service providers, and enhance their competitive advantage by relying on software platforms and full-link services.