On October 15, Haiguang Information (688041.SH) released the first A-share semiconductor earnings report for Q3 2025. The company achieved an operating revenue of 9.49 billion yuan in the first three quarters, a year-on-year increase of 54.65%; the net profit attributable to the parent company was 1.961 billion yuan, a year-on-year growth of 28.56%. In Q3 alone, Haiguang Information reported an operating revenue of 4.026 billion yuan, up 69.6% year-on-year; the net profit attributable to the parent company reached 760 million yuan, reflecting a 13.04% year-on-year increase, setting a new historical record. Discussing the reasons for the performance growth, Haiguang indicated that the company accelerated client adoption by deepening collaboration with OEMs and ecological partners in key industries and sectors, which significantly bolstered sales of its high-end processor products. Beyond Haiguang, several semiconductor listed companies are also expected to report earnings growth in Q3 this year. According to statistics from the “Star Market Daily,” seven semiconductor companies have disclosed performance forecasts so far, with the results illustrated in the following table. Among them, Changchuan Technology posted the highest growth rate. The company attributed the performance change to the continuing increase in market demand for semiconductors during the reporting period, leading to robust client demand and abundant product orders, which have substantially boosted sales revenue and profits. Guotai Junan Securities pointed out that in 2025, the CIS and memory testing machines will gradually enter a ramp-up phase, and it is expected that the net profit attributable to the parent company of Changchuan Technology will exceed market expectations for the full year. Within the aforementioned companies, five are engaged in the chip design sector, including Tailin Microelectronics and Juxin Technology. In their announcements, several companies unanimously mentioned the impact of the end-side AI market on earnings: Tailin Microelectronics stated that during the reporting period, shipments of their end-side AI chips exceeded expectations, with rapid expansion in overseas markets leading to a revenue increase of approximately 30% year-on-year. Ruixin Microelectronics reported that their RK3588 products have penetrated automotive electronics, machine vision, industrial applications, and robotic markets, resulting in rapid revenue growth and a significant rise in net profit. Juxin Technology noted that the penetration rate of end-side AI processor chips has surged, with related sales revenue experiencing several-fold increases. Currently, large models and generative AI are quickly transitioning from the cloud to consumer endpoints. Founder Securities noted that as model miniaturization technology matures and hardware upgrades in computing power, storage, and connectivity continue to progress, an explosion in AI terminal usage is imminent. Although domestic manufacturers of mainstream end-side computing power chips like AI smartphones and AIPC have started relatively late, domestic firms have significant potential in end-side AIoT computing, storage technologies, and wireless connectivity chips, promising opportunities for catching up. CITIC Securities projected that with the consumer electronics peak season and intensive releases of new end-side AI products in the second half of this year and next year, domestic and foreign manufacturers’ positive capital expenditures related to AI will drive a favorable industry outlook. Coupled with innovations, upgrades, and domestic alternatives, the industry cycle is expected to continue its upward trajectory, underscoring overall investment value. From an investment perspective, this institution highlighted that current overseas resources for advanced manufacturing and packaging capacity are scarce, while the domestic computing power industry chain is still striving to catch up, with key areas facing a state of “strong demand but weak supply.” With a solid foundation already established for domestic semiconductor localization, there remains significant room for improvement in advanced processes, storage, packaging, core equipment materials, and EDA software. Emphasis is advised on the growth in AI demand and advancements in localization.