After market close on October 27, Ingenic (300223.SZ) released its 2026 Q3 report: revenue for the first three quarters reached RMB 7.102 billion, up 106.66% year-on-year; net profit attributable to shareholders was RMB 2.616 billion, up 922.95%; and non-GAAP net profit was RMB 2.595 billion, up 1,188.99%. The third quarter alone was even steeper — revenue of RMB 3.145 billion, up 164.85% year-on-year; net profit of RMB 1.419 billion, up 2,595.49%; and non-GAAP net profit of RMB 1.429 billion, up 3,436.84%. Revenue, net profit, and non-GAAP net profit also grew 30.35%, 99.65%, and 105.62% quarter-on-quarter, respectively.
Looking at the business structure, this Fabless IC design company runs three product lines: memory chips covering DRAM, SRAM, and Flash; computing chips including smart-vision SoCs, embedded MPUs, and AI MCUs; and analog chips spanning LED drivers, touch sensing, DC/DC, and network-transmission chips such as LIN, CAN, and GreenPHY. In the first three quarters, memory revenue was RMB 4.916 billion, up 130.73% year-on-year with a 56.53% gross margin — tight DRAM supply kept pushing prices up, and Flash demand kept growing in emerging fields such as AI servers. Computing revenue was RMB 1.752 billion, up 92.89% with a gross margin as high as 64.04%, and analog revenue was RMB 418 million, up 11.70% with a 49.28% gross margin. Net operating cash flow was RMB 1.896 billion, up 271.53%. On the sell side, Huaxi Securities and China Post Securities project full-year 2026 revenue of RMB 9.95 billion and RMB 9.5 billion respectively.
From the perspective of the industry landscape, Ingenic’s single-quarter surge is a textbook sample of “the memory supercycle resonating with on-device computing.” On one hand, the price-upcycle driven by the DRAM supply gap directly lifted the memory line into simultaneous volume and price growth; on the other, the 64% gross margin of computing chips shows that the profitability structure of domestic computing chips — smart-vision SoCs, embedded MPUs, and AI MCUs — has improved markedly, as on-device AI inference demand creates a second growth curve for computing silicon. What makes the picture intriguing is the expectation gap: the stock fell a cumulative 26.75% from October 1 to 27, while the institutional forecast range for full-year revenue already stands at RMB 9.5–9.95 billion, and the RMB 7.102 billion booked in the first three quarters covers more than 70% of those forecasts. The contrast between delivered results and a falling share price reflects the market’s disagreement over how long the memory cycle can last.
For Junyi Microelectronics, Ingenic’s report confirms the buoyancy of the embedded computing ecosystem. Ingenic’s embedded MPUs and AI MCUs belong to the same on-device computing-chip ecosystem as Junyi’s core boards J-3568/J-3576/J-3588 and K-series edge-intelligent motherboards — the high prosperity of computing chips indicates robust downstream demand for end products and boards, corroborating the embedded computing track where Junyi operates. The memory price-upcycle, meanwhile, is a reminder that hardware makers need stronger material-management capabilities: through standardized core-board design and BOM stocking strategies, Junyi hedges the impact of DRAM/NAND price swings on whole-machine costs. The continued validation of on-device AI demand also underpins the long-term volume logic of the edge-computing hosts 5G-EDGE-3399Pro/AI-BOX-3588 and the industrial mini PC IPC-i5-1235U in industrial scenarios. In the end, volume and price signals from the supply chain always transmit into the product definition and stocking rhythm of end-product makers.
Source: https://baijiahao.baidu.com/s?id=1880253296398846329 (2026-10-28)