In this edition of the Daily Sci-Tech Innovation column, we turn our attention to a relatively low-profile company listed on the SSE STAR Market in the field of computer vision. It does not manufacture smartphones or automobiles, nor does it sell cameras. However, the algorithms powering your smartphone photography and the fatigue monitoring system in your car's cabin are highly likely developed by this company.
ArcSoft (688088) has just released its 2026 semi-annual report: revenue reached CNY 440 million, a year-on-year increase of 7.19%; net profit attributable to shareholders was CNY 77.1384 million, a year-on-year decline of 12.87%; and net profit excluding non-recurring gains and losses dropped by nearly 22%. With revenue growing while profits shrinking, where does this "scissors gap" come from? Where exactly lies the moat of this company? In this article, we will try to figure out the numbers.
01. Not Selling Hardware, but the "Underlying Core Capabilities" for Smartphone Photography
What ArcSoft does can be summarized in one sentence: providing visual AI (Artificial Intelligence) algorithms for smart devices. Its business model does not involve selling hardware or modules; instead, it "installs" its independently developed visual algorithms into customers' terminal products through licensing, charging licensing fees on a per-project basis.
The defining feature of this model is an exceptionally high gross profit margin. In the first half of 2026, the company's overall gross margin reached 89.03%, with the gross margin for its mobile smart terminal visual solutions hitting as high as 91.33%. For every CNY 1 of revenue, nearly CNY 0.9 is gross profit—this is almost unimaginable in the hardware supply chain, but it is the norm in a pure software licensing business.
Mobile smart terminals constitute ArcSoft's core business. In the first half of the year, revenue from this segment reached CNY 359 million, a year-on-year increase of 5.78%, accounting for over 80% of total revenue. The client list includes major global Android smartphone manufacturers such as Samsung, Xiaomi, OPPO, and vivo. In this field, ArcSoft is a leading global supplier of camera algorithms for Android smartphones, holding a remarkably solid position.
However, the smartphone market itself is not having an easy time. According to statistics from third-party agencies, global smartphone shipments in the second quarter of 2026 declined by 11% year-on-year. Rising prices for memory chips have further squeezed the profit margins of mid-to-low-end models, and customers have imposed stricter requirements on the memory footprint, model size, and operational efficiency of algorithms. ArcSoft's smartphone business growth rate has fallen from double digits in past years to single digits, which is more a result of industry-level pressure rather than an issue with the company's own competitiveness.
02. 53% of Revenue Poured into R&D: Not "Burning Cash," but "Making a Bet"
The core reason for the profit decline is stated very clearly in the semi-annual report: R&D expenses reached CNY 235 million, a year-on-year increase of 18.3%, and the R&D expense ratio climbed to 53.41%, up 5.02 percentage points year-on-year. Looking solely at the second quarter, the R&D expense ratio even surged to 57.7%.
What does this figure mean in the context of A-share technology companies? For most A-share software companies, the R&D expense ratio is roughly between 15% and 25%, and exceeding 30% is generally considered "heavy R&D." ArcSoft investing over half of its revenue into R&D indicates that the company's strategic choice at this current stage is very clear: trading profits for the future.
Where is the money being spent? In three directions. First, the continuous iteration of the "TurboFusion" technology for smartphones, aiming to further improve the four scenarios most easily perceived by consumers: portrait, telephoto, ultra-clear image quality, and video tonality. Second, in-vehicle AI vision, where DMS (Driver Monitoring System) and OMS (Occupant Monitoring System) products are moving from designated projects to mass production, while the low-computing-power platform SouthLake is expected to begin large-scale shipments in the second half of the year. Third, emerging tracks, where AI glasses already have multiple projects in mass production and shipment, and the core visual algorithm engine for robots has completed adaptation to mainstream technology ecosystems such as NVIDIA, Intel, Qualcomm, and Rockchip, and has been delivered to leading robot customers.
The number of R&D personnel increased from 619 at the end of 2025 to 659, an addition of 40 people, leading to a corresponding increase in salary and social security expenditures. This is a "front-loaded cost"—the products are still in the validation and ramp-up stages, and revenue has not yet been realized on a large scale, but human resources and R&D investments have already been incurred. In its research report, Huatai Securities gave a "buy" rating under the title "Profit End Under Pressure Due to Phased Industry Impacts," while also expressing long-term optimism about the steady growth of mobile smart terminals and in-vehicle AI businesses.
03. In-Vehicle Business: The Next Growth Engine Worth Watching
If ArcSoft is regarded merely as a "smartphone algorithm company," it would certainly be underestimated. Revenue from in-vehicle AI vision solutions reached CNY72 million in the first half of the year, a year-on-year increase of 11.71%. Although the scale is still small, the growth rate has already surpassed that of the smartphone business.
More crucial is the elevation of its market position. According to Gasgoo data, among domestic passenger car DMS (Driver Monitoring System) algorithm suppliers from January to November 2024, ArcSoft ranked third with a market share of approximately 13%, trailing only SenseTime Auto and Tesla. Its share increased significantly from 8.3% in the same period of 2023 (the figure for January to April 2025 was 10.7%, still ranking in the top three). Considering that DMS products themselves have a low unit value but extremely high technical thresholds, this is a market where share continues to concentrate among top players. Relying on its algorithm reuse capabilities, technical synergy with sensor manufacturers like OmniVision and Sony, and deep adaptation with cabin chip platforms such as Qualcomm, ArcSoft is gradually capturing market share.
In terms of customer progress, DMS compliance products have been designated by multiple clients including Voyah, Hyundai, and Chery. OMS has added mainstream carmakers such as Geely, Xiaomi, and Great Wall. The intelligent parking assistance system has completed initial delivery on the SouthLake platform, entering the commercialization stage. Domestic business growth is limited due to vehicle price cuts and cost pressures, but overseas in-cabin business is progressing positively, and the company expects European projects to see volume growth in the second half of the year.
04. Investment Logic and Risks: A Game of "Time Lag"
The starting point of the investment logic is a clear time lag. The core contradiction ArcSoft currently faces is that R&D investments and early layouts for emerging businesses have already occurred, but the realization of revenue still requires a ramp-up cycle of one to two years. New directions such as AI glasses, robot vision, and intelligent commercial photography (PSAI) are mostly still in the stage of "moving from technical validation to mass production." If these businesses see concentrated volume growth around 2027, the company's revenue structure and profit quality will undergo substantive changes; if the volume growth pace falls short of expectations, high R&D investments will continue to suppress profit performance.
The second logic lies in the scarcity of the business model. In a pure algorithm licensing business, once it enters a customer's supply chain, the switching costs are very high, resulting in strong customer stickiness. The proportion of sales to ArcSoft's top five customers to total sales has long been maintained at around 60% (60.14% in 2024 and 58.69% in 2025). This is both a concentration risk and evidence of its moat.
Risks are equally impossible to ignore. The most direct one is the downstream prosperity risk. The smartphone business accounts for 80% of revenue, and fluctuations in global smartphone shipments will directly transmit to ArcSoft's revenue end. Although the in-vehicle business is growing fast, the price-cutting pressure from domestic carmakers is being transmitted to upstream algorithm suppliers. The unit price of products like DMS is inherently low, leaving little room for price wars. Furthermore, the expansion of new businesses carries uncertainties; the market pace of AI glasses and robot vision is not entirely controlled by ArcSoft, and customers' product definitions and marketing effectiveness will directly affect the procurement volume of algorithms. Additionally, the net cash flow generated from operating activities in the first half of the year was CNY -10 million, shifting from a net inflow in the same period last year to a net outflow. The divergence between profit and cash flow is also worth tracking.
Therefore, the "increased revenue but decreased profit" figure in the semi-annual report is better described as a normal stumble during a gear-shifting period for a technology-driven company, rather than a sudden deterioration in performance. What is truly worth tracking is not the profit figures for the next quarter, but the mass production orders for AI glasses, the shipment pace of the SouthLake platform, and when the turning point will occur for robot vision customers to shift from "technical cooperation" to "batch procurement."
- End -