Author | ZeR0 Editor | Moying
Xindongxi reported on October 26 that yesterday, Jiangsu Suzhou-based General Purpose GPU (GPGPU) company Denlin Technology passed the listing hearing of the Hong Kong Stock Exchange.
Founded in November 2017, Denlin Technology relies on its proprietary GPU chip architecture, the GPU+* (Minsky) architecture (which integrates GPGPU cores, AI acceleration units, and high-bandwidth on-chip memory subsystems into a single chip via intra-chip heterogeneity). The company targets Enterprise AI as its core market while striving to expand into public cloud markets and innovative fields such as robotics. Its senior independent investors include China Internet Investment Fund, Qualcomm China, Sino-Singapore Venture Capital, and the Integrated Circuit Growth Fund, among others. Other industrial investors include Hikvision's subsidiaries Hik Fund and Haina Yuzhi, Zhonghe Auto, and Oriental Semiconductor.
According to Frost & Sullivan data, based on revenue in 2025, Denlin Technology ranked top ten in the Chinese GPGPU market, with a market share of less than 1.0%; it was the third-largest company in the Chinese Enterprise AI GPGPU market, with a market share of 1.0%.
01.Joint Entrepreneurship by Multiple Tsinghua Alumni, Valuation Reached CNY 7.2 Billion After Series D++ Round
According to the prospectus, Denlin Technology was founded in November 2017 by Li Jianwen, Wang Chenhui, Wang Ping, and the late Wang Zhenyu, among others. Li Jianwen holds bachelor's and master's degrees from Tsinghua University in lithography and process engineering, currently serving as Chief Strategy Officer. Wang Chenhui holds a bachelor's degree in Computer Science and Technology from Shanghai University, serving as Chairman of the Board, Executive Director, and CEO. Wang Ping graduated from Tsinghua University with a major in Microelectronics, currently serving as Chief Innovation Officer.
Li Jianwen, Wang Chenhui, Wang Ping, Denglin Consulting, Denlin Xingyi, Denlin Longgui, Denlin Wanhua, and Lingxiao Ventures constitute the controlling shareholders under the Listing Rules and are parties to the concerted action arrangement. From March 2018 to December 2025, Denlin Technology completed multiple rounds of financing, raising over CNY 1.5 billion cumulatively, with a post-Series D++ valuation of CNY 7.19 billion.
02.Revenue Reached CNY 200 Million in First Four Months of This Year, Still Not Yet Profitable
Denlin Technology has developed multi-generation GPGPU product portfolios covering enterprise data center products and edge inference products. These products aim to provide foundational computing power for enterprise AI applications across various deployment environments (including edge nodes, public clouds, private clouds, and terminal devices), built upon an original GPU+ (Minsky) architecture supported by a unified Hamming software stack.
Taking enterprise private cloud applications as an example, the Goldwasser series is specifically designed for large-scale enterprise-level AI workloads, offering competitive performance density, energy efficiency, and deployment flexibility. By supporting FP32, TF32, FP16, BF16, and INT8 calculation precisions, the boards can execute multi-modal tasks in parallel within a unified infrastructure, such as model training, inference, image rendering, and analysis. These capabilities make its platform highly suitable for enterprise AI computing clusters and data centers. In 2023, 2024, 2025, and January-April 2026, Denlin Technology's revenues were CNY 28 million, CNY 83 million, CNY 171 million, and CNY 200 million respectively; net profits were -CNY 332 million, -CNY 297 million, -CNY 281 million, and -CNY 16 million respectively; R&D expenses were -CNY 258 million, -CNY 248 million, -CNY 238 million, and -CNY 65 million respectively.
During the same period, its adjusted net profits were -CNY 278 million, -CNY 224 million, -CNY 204 million, and -CNY 21 million respectively.
By business segment, from 2023 to the end of April 2026, over 67% of the company's revenue came from the GPGPU business, with the remainder coming from chip design services and IP licensing.
Its GPGPU products have been deployed in six key application scenarios: smart cities, system integration, enterprise private clouds, industrial quality inspection, education, and industry-specific large models.
03.GPGPU Sales Exceeded 10,000 Units in First Four Months of This Year, Fully Autonomous Core IP Portfolio Developed
An overview of the key performance specifications of Denlin Technology's main product series includes:
- The Goldwasser I product based on Minsky I offers up to 256 TOPS (INT8), 64 TFLOPS (FP16), and 16 TFLOPS (FP32) computing power, equipped with 8GB to 64GB memory, memory bandwidth up to 239 GB/s, and maximum power consumption of 55W, primarily used for inference tasks in smart city scenarios.
- The Goldwasser II product based on Minsky II offers up to 512 TOPS (INT8) and 256 TFLOPS (FP16) computing power, equipped with 32GB to 128GB memory, memory bandwidth up to 192 GB/s, and maximum power consumption of 95W, supporting inference and training in smart cities, intelligent computing centers, and other scenarios.
- The KS series products based on Minsky III offer up to 560 TOPS (INT8) and 280 TFLOPS (FP16) computing power, equipped with 8GB to 128GB memory, memory bandwidth up to 819.2 GB/s, and maximum power consumption of 130W, simultaneously supporting inference and training.
Shipments of the KS series increased in the second half of 2025 and are expected to continue growing throughout 2026.
Second-phase products are designed to serve enterprise AI customers and large model training and inference workloads for intelligent computing centers, expected to deliver up to 2000 TFLOPS of FP8 computing power. Target customers include large intelligent computing center operators and public cloud providers. Total sales volume of Denlin Technology's GPGPUs increased from 4,199 units in 2023 to 5,295 units in 2024, further surging to 12,144 units in 2025, and reaching 10,344 units in the first four months of 2026.
The company expects to launch the first phase and second phase of Minsky IV products in December 2026 and the second half of 2027 respectively, further expanding the GPU+ platform into high-performance domains. The first phase will continue to use LPDDR-based on-chip memory subsystems, while the second phase will integrate HBM3e or other advanced memory technologies to meet the demand for extreme throughput and memory in public cloud and intelligent computing center scenarios. Based on its unique GPU+ (Minsky) architecture, Denlin Technology has developed a structured and fully autonomous core intellectual property (IP) portfolio, covering a wide range of computing and acceleration functions. Unlike IP suppliers such as Arm or Imagination Technologies (whose GPU IPs are typically designed for mobile applications and lack the performance required for high-end cloud workloads), the company focuses on cloud-scale, high-performance applications that require massive architectural integration and optimization.
To complement its self-developed hardware, the company has built a fully integrated software platform—the Hamming software stack.
As of April 30, 2026, nearly 70% of its employees were R&D personnel, most of whom have over ten years of industry experience.
04.Top Five Customers Account for Over 83% of Revenue, Most Customers Are Private Enterprises
In 2023, 2024, 2025, and January-April 2026, the revenue contribution from Denlin Technology's top five customers accounted for 98.4%, 85.1%, 83.0%, and 90.4% of total revenue respectively, while the largest customer's revenue accounted for 37.9%, 25.7%, 29.4%, and 38.0% of total revenue during the same periods.
During the same periods, purchases from the top five suppliers accounted for 83.1%, 65.2%, 56.9%, and 76.3% of total procurement value respectively, while purchases from the largest supplier accounted for 59.5%, 31.1%, 31.2%, and 21.0% of total procurement value respectively.
Several entities are both customers and suppliers of Denlin Technology. For instance, Denlin Technology procured customized chip design services from Supplier F and previously collaborated with Supplier F on early product designs targeting overseas process nodes; it procured AI edge computing equipment from Customer D and server equipment from Customer L. During the past record periods, most customers of Denlin Technology's GPGPU products were private enterprises, with the rest being listed companies. To support continued capacity expansion, Denlin Technology has placed orders to secure core wafer capacity resources equivalent to approximately 1,400 wafers, while expanding packaging, testing capabilities, and PCB assembly capacity through third-party partners. Additionally, Denlin Technology is negotiating with other potential clients, with intended contract amounts totaling approximately CNY 200 million.
05.Conclusion: Three Major Drivers Bring Opportunities to the GPGPU Market
Denlin Technology believes its market opportunities stem from three major drivers:
First, AI inference demand is shifting from public clouds to the enterprise side. Enterprises have rapidly growing rigid demands for low-latency, high-energy-efficiency GPGPUs (such as for real-time video analysis and industrial defect detection), which aligns well with the company's technical accumulation in heterogeneous architecture optimization (such as improving computing density and controlling energy consumption);
Second, emerging fields (such as industrial simulation and the metaverse) release computing power demands, driving the continuous expansion of GPGPU solutions at edge nodes;
Third, national policies strongly support the development and application of domestic GPGPUs, including the "Eastern Data Western Computing" project optimizing computing layout and the IT Application Innovation (Xinchuang) procurement policies, accelerating the large-scale penetration of domestic GPUs in key industries. In response, Denlin Technology captures long-term opportunities arising from the collaborative development of industry-specific large models and AI agents through solutions encompassing software-hardware integration.