Revenue Growth Without Profit Increase: The Awkward Report Card of "Autonomous Driving Chip First Stock" Black Sesame Intelligence

(Source: Meicai.com)

Black Sesame Intelligence’s 2025: A Mix of Joy and Sorrow, Revenue Breaks 800 Million Yuan, Losses Reach 1.48 Billion Yuan

Text / Daily Financial Report Chu Feng

With the halo of being the “First Domestic Smart Driving Chip Company,” Black Sesame Intelligence delivered a mixed report card for 2025: annual revenue is expected to surpass 800 million yuan, a nearly 70% increase year-over-year, but net losses will also reach 1.48 billion yuan.

Behind these contradictions lie the company’s challenges: insufficient computing power of main products leading to weak bargaining power, a single customer structure creating growth ceilings, and high R&D investments continuously draining cash flow.

As diversification into robotics and acquisitions like Ezviz Electronics become new growth strategies, market attention is focused on whether these measures are effective solutions or just another battlefield for resource dispersion. For Black Sesame Intelligence, time may be running out.

High Revenue Growth Cannot Mask Profitability Shortfalls

In 2025, Black Sesame Intelligence posted impressive revenue growth, with total revenue expected to exceed 800 million yuan, a growth of over 68.7% year-over-year. This growth mainly relies on increased mass production of high-end smart driving chips in the passenger car market, large-scale shipments of L2-L3 commercial vehicles and L4 level driverless logistics, and income from new robotics business.

In the first half of 2025, the company’s revenue also grew by 40.4% year-over-year to 253 million yuan, achieving scale production in multiple models with partners like Geely and Dongfeng.

However, behind this revenue growth lies an unavoidable profitability dilemma. In the first half of 2025, the company’s net loss reached 762 million yuan, a significant decrease of 169% compared to the previous year. The full-year net loss is expected not to exceed 1.48 billion yuan, whereas in 2024, the company still achieved a net profit of 313 million yuan.

In fact, the profit in 2024 was not from core operations but from fair value changes in financial instruments issued to investors. Excluding this factor, the company’s operating profit has been negative for a long time.

Currently, Black Sesame Intelligence’s main business remains focused on autonomous driving products and solutions, and intelligent imaging solutions. In the first half of 2025, these two segments achieved revenues of 237 million yuan and 16 million yuan, respectively, with year-over-year increases of 41.5% and 24.8%. Gross profit margins were 20.9% and 82.4%, both showing declines compared to the same period last year.

The company attributes the decline in gross margin to increased hardware component and labor costs due to expansion into new application scenarios. This “price-to-volume” expansion strategy has driven revenue growth but has directly squeezed profit margins.

Meanwhile, high R&D costs continue to deplete the company’s already tight cash flow. In the first half of 2025, R&D expenditure was 618 million yuan, accounting for 244% of revenue. Even though this was lower than the previous year, it remains a significant expense.

Factors such as stock-based compensation and reduced fair value gains on financial instruments further increase the company’s loss pressure. As of June 2025, the company’s cash and cash equivalents stood at 1.966 billion yuan, with cash flow under pressure becoming a real issue.

Main Product Competitiveness Faces Challenges

In the highly technical field of smart driving chips, product computing power and customer recognition are fundamental for survival. In the domestic high-performance smart driving chip market, Horizon Robotics holds a 28.65% market share, ranking first, while Black Sesame Intelligence has not yet entered the top five.

Currently, Black Sesame’s flagship chip A1000 has a computing power of 58 TOPS, suitable for L2+/L3 assisted driving. Compared to competitors—Horizon’s Journey 6M with 128 TOPS and Nvidia’s Orin-X with 254 TOPS—Black Sesame’s chip is significantly lagging. Facing mainstream automakers’ demand for higher computing power, Black Sesame finds itself in an awkward middle ground.

Although the next-generation A2000 chip has been approved for global sales, it uses 7nm process technology, offers over 250 TOPS AI computing power, and supports full-scene urban NOA applications, it is still in the development and validation stage and has not yet been officially adopted for mass production by any automaker. This means that at a time when the market most needs new products, the company can only rely on its previous generation.

The fragility of customer structure further amplifies the product competitiveness issue. Despite claims of cooperation with leading clients like Geely, BYD, Dongfeng, and FAW, most of their models are not bestsellers.

For example, Geely’s Galaxy Star Shine sold over 10,000 units in August, but Galaxy E’s August sales have fallen to a few hundred units. In contrast, Horizon’s customer list includes not only BYD and Li Auto but also major orders from Volkswagen China.

More critically, some orders come from Tier 1 suppliers like Bosch and ZF, rather than directly from OEMs, which weakens their sensitivity to end-user demand and bargaining power. In the first half of 2025, sales costs surged 111.2% year-over-year to 190 million yuan, accounting for 76% of revenue, indicating a passive position in commercial negotiations.

It is also worth noting that Black Sesame Intelligence completed a HKD 631 million private placement on March 10, with Wuji Capital as the investor. The funds will be mainly invested in developing new high-performance chips, including high-computing chips for autonomous driving and robotics, edge AI chips for terminal devices, and overseas R&D centers, providing strong support for the company’s long-term technological layout and enhancing product competitiveness.

Diversification as a Solution to Urgent Needs?

Faced with sluggish growth in core business, Black Sesame Intelligence is trying to find new growth points through diversification. Robotics has become a key focus. In November 2025, the company launched the SesameX full-stack computing platform, announcing entry into humanoid robotics, and revealed initial partners including Yunshen and Fourier Intelligence.

By the end of 2025, the company acquired a 60% stake in Ezviz Electronics, a low-power, high-cost-performance AI chip company, for 478 million yuan, aiming to fill technical gaps in high-performance AI chips and form a product matrix covering high, mid, and low-end segments.

From a technological transfer perspective, this strategy is reasonable. Both autonomous driving and robotics require solving complex environment perception and rapid decision-making. The experience accumulated in automotive-grade chips has some reuse value. In the first half of 2025, robotics contributed to revenue.

However, humanoid robotics is still in early exploration, with fragmented chip demands, undefined standards, and a 5-10 year timeline for commercialization. Large-scale investment at this stage is high risk, and the significant differences in computing architecture and power consumption requirements between driving and robotics chips mean core technologies are difficult to reuse, risking resource dispersion.

Expanding into overseas markets also faces challenges. Although the Huashan A2000 chip has been approved for global sales by U.S. authorities, and overseas vehicle models and quantities hit record highs in the first half of 2025, ongoing global tariff conflicts and policy barriers against Chinese chips in Europe and America mean actual business results remain uncertain. As of press time, the company has not disclosed specific overseas revenue data.

For Black Sesame Intelligence, time may be limited. While high revenue growth is encouraging, it cannot hide the limitations of a “burn money for growth” model. Whether the company can successfully launch the next-generation A2000 chip and open up the market before cash runs out will be crucial to reversing the situation. Ultimately, the path to breakthrough depends on the core competitiveness of its products.