Hello! I'm your financial analyst friend. Today, we're going to talk about a company called Suiyuan Technology.
In simple terms, it's the last of the "four major players" in the domestic AI chip industry to go public. Its story is quite remarkable: two veterans with many years of experience at AMD (a global leading chip company), along with a strong background from Tsinghua University and substantial funding from Tencent, started from a small office of less than 100 square meters in Zhangjiang, Shanghai. In just 8 years, they built the company into one with a market value of over 170 billion yuan.
However, behind its glamorous success, there are several challenges and concerns. Today, I'll break down the company's situation in plain language to help you understand its true potential and whether it's worth your attention.
---
1. The “Big Profit” on the First Day of Listing: How Was the Money Earned?
First, let's look at how popular it was on the day of its listing.
On September 11, 2026, Suiyuan Technology went public on the STAR Market. The issue price was 142.18 yuan, and the stock opened at 410 yuan, closing at 397 yuan. This means that if you were lucky enough to get a share (1,000 shares), you could have made a profit of 128,000 yuan. In the stock market, this is called a “big profit” opportunity.
Why did everyone rush to buy?
1. Scarcity: It was the last of the “four major players” to go public, so the market was already excited about the previous companies (Cambricon,摩尔Thread, and Muxi), making Suiyuan the focus.
2. High Growth Expectations: Although it wasn't yet profitable, its revenue grew rapidly, from 300 million yuan in 2023 to 990 million yuan in 2025, more than tripling in just three years. This kind of growth is very rare in the chip industry.
3. Tencent’s Support: Tencent is its largest shareholder and customer. Having a giant like Tencent backing it is like having a huge “credit check” in the capital market.
In plain language: People bought Suiyuan not because of its current profits (it was still losing money) but because of its potential for future growth and the security that comes with having Tencent as a major customer.
---
2. The Founders (Zhao Lidong) and Tencent (The Investor): Is This Combination Stable?
Suiyuan Technology has two key figures:
- Zhao Lidong: Graduated from the Class of 1985 in the Department of Electronic Engineering at Tsinghua University (a legendary class that produced many industry leaders), he worked at AMD in Silicon Valley for over 20 years as a senior product engineer and now also serves as the chairman and secretary of the board.
- Zhang Yalin: Another former AMD colleague, responsible for technology and products.
How do they divide the work?
- Zhang Yalin manages technology, research and development, and personnel.
- Zhao Lidong handles finance, strategy, and external relations.
Why is this combination important?
The chip industry is both costly and has a long development cycle. Many young entrepreneurs know technology but don’t know how to raise funds, manage supply chains, or navigate complex business environments. Zhao Lidong’s experience at AMD and Tsinghua Group gives him valuable industry resources.
Tencent’s Role:
Tencent has invested in Suiyuan for several rounds and is not only a shareholder but also its largest customer. In 2025, 80% of Suiyuan’s revenue came from Tencent. The “voice-to-text” function in WeChat uses Suiyuan’s chips.
In plain language: It’s like running a restaurant where the chef (Zhang Yalin) is skilled, the owner (Zhao Lidong) knows how to manage, and the main customer (Tencent) not only buys the food but also helps with renovations and promotion. This combination of technology, capital, and orders is crucial for Suiyuan’s success.
---
3. Focusing on Specialized Skills: What Are the Pros and Cons?
This is a crucial point to understand Suiyuan.
There are two main types of AI chips:
1. General-Purpose (GPGPU): Like NVIDIA’s chips, which can do everything from drawing to complex calculations and training large models, with a large ecosystem (CUDA), but they are expensive and power-consuming.
2. Specialized (DSA): Such as Suiyuan’s, Huawei’s Ascend, and Cambricon’s, which are designed for efficient AI computing. They remove unnecessary components and focus on the core computing tasks.
Suiyuan focuses on the DSA approach.
- Pros: High cost-effectiveness. For companies like Tencent, which process large amounts of voice and video data daily, saving power means saving money, and higher efficiency means more profit.
- Cons:
- Ecosystem Limitations: Their chips are not compatible with NVIDIA’s CUDA ecosystem, so reprogramming is required when switching from NVIDIA to Suiyuan.
- Weaknesses: They are good at inference (quick responses to AI queries) but not as strong at training (learning from large datasets). Training requires more versatile computing power, where Suiyuan falls short compared to NVIDIA.
In plain language: NVIDIA is like an “all-rounder” athlete, good at various tasks but not at the highest level, and expensive. Suiyuan is like a “sprint champion” fast but not as versatile. It mainly makes money from inference tasks for large companies.
---
4. How Good Is Its Profitability? Why Is the Valuation So High?
Let’s look at its financial performance:
- Still Losing Money, but Losing Less: Revenue grew from 300 million yuan in 2023 to 990 million yuan in 2025, and in the first half of 2026, it already exceeded the previous year’s total.
- High Research Costs: Research and development expenses were 1.135 billion yuan, more than double its revenue, which is common in the chip industry. Without investing in R&D, companies fall behind.
- Gross Margin: Rising from 22.6% to 30.6% to 31.8%, indicating maturing products and better cost control.
- High Valuation: A market value of 170.8 billion yuan and a price-earnings ratio (PSR) of 94, meaning the market is willing to pay 94 yuan for every 1 yuan in revenue. This is higher than Cambricon’s 68 times,摩尔Thread’s 66 times, and Muxi’s 95 times.
- Reason for High Valuation: The market sees inference as the future trend. While AI was initially used for training large models (like GPT-4), it’s now widely used in mobile devices, computers, and cars for tasks like real-time translation and image generation, which require efficient chips. Suiyuan’s focus on inference aligns with this trend.
In plain language: Suiyuan is in a phase of investing heavily to build its future. It’s not yet profitable, but its rapid growth (81% revenue increase) and its focus on inference make it a target for investors who hope it will become a profitable business.
---
5. Major Risks: Heavy Dependence on Tencent and Multiple Product Development Challenges
Here are some potential issues:
Heavy Customer Dependence: 80% of its revenue comes from Tencent, which could be a major risk if Tencent develops its own chips or switches suppliers.
- Challenge: Suiyuan needs to prove it can sell to other customers besides Tencent and replicate its technology.
- Training Chip Weakness: Although the inference market is large, the training market is more lucrative. Its fourth-generation chip, L600, hasn’t been mass-produced yet. If it lags behind in performance, Suiyuan may struggle in the high-end market.
- Multiple Product Development: Suiyuan is working on three generations of chips simultaneously, which can be challenging and resource-intensive. Technological advancements can quickly render current products obsolete.
- Fierce Competition: It faces strong domestic and international competitors, including Huawei Ascend, Cambricon, and MooreThread, as well as the dominant NVIDIA.
In plain language: Suiyuan is at a critical juncture, relying on Tencent and managing multiple product development projects simultaneously. It must balance growth and risk management.
---
Summary: What Is Suiyuan Technology’s True Potential?
Strengths:
- Strong Team: Veteran AMD professionals with Tsinghua backgrounds and Tencent’s support.
- Accurate Focus: Specialized DSA chips that meet current AI trends.
- Rapid Growth: Tripled revenue in three years, with strong momentum in the first half of 2026.
- Stable Customer: Tencent provides stable orders and market validation.
Weaknesses/Risks:
- High Dependence: 80% revenue from Tencent, increasing vulnerability to changes in Tencent’s strategy.
- Technological Challenges: Weakness in training chips, affecting its competitiveness in high-end markets.
- High Valuation: The high valuation may not be sustainable if performance doesn’t meet expectations.
- Fierce Competition: Facing strong competitors both domestically and internationally.
Advice for Investors:
If you’re optimistic about the long-term prospects of AI inference chips and believe Suiyuan can reduce its dependence on Tencent, it’s worth watching. However, its stock price can be volatile due to market sentiment and performance gaps.
In one sentence: Suiyuan Technology is a company with high growth potential and a high valuation, but it also faces significant risks. It represents the challenges of China’s emerging tech startups: a promising start but a long road with many hurdles to overcome. Whether it can become a “Chinese NVIDIA” depends on its ability to break away from its reliance on Tencent and continue to innovate.