On August 18, 2026, US AI chip startup Cerebras Systems (stock code: CRS) officially listed on the Nasdaq Exchange, with an issue price of $25 per share and total fundraising of approximately $1 billion. On the first day of trading, CRS stock price jumped to $62.5 at the open, up 150% from the issue price, closing at $58.3, a gain of 133.2%, with a total market capitalization of $5.83 billion, becoming one of the most outstanding companies among US AI new stocks this year.
Cerebras Systems was founded in 2015, headquartered in California, focusing on developing high-performance chips for AI training. Its core product is the "WSE-2" (Wafer-Scale Engine 2), the world's first AI training chip using wafer-scale integration technology. A single chip integrates 1.2 trillion transistors, has 85 million cores, and delivers 20 PFLOPS (20 trillion floating-point operations per second), more than 10 times that of traditional GPU chips. This chip is designed specifically for large-scale AI model training, significantly shortening training time and reducing energy consumption, making it an important innovation amid the surge in AI computing power demand.
1. Technical Advantages: Wafer-Scale Integration and Specialized Architecture
Cerebras' core technical breakthrough lies in "wafer-scale integration" (Wafer-Scale Integration), which means no longer cutting chips into small chips but using the entire wafer as a single chip, thus eliminating inter-chip interconnect latency and significantly increasing computing power density. Compared to traditional GPU chips, WSE-2's computing power density is 100 times higher, and energy consumption is 5 times lower, giving Cerebras' chips a clear advantage in training large AI models (such as GPT-4, DALL-E 3).
Additionally, Cerebras adopts a "specialized architecture" (Specialized Architecture) optimized for the specific needs of AI training. For example, its chip is equipped with high-speed memory (HBM), with a memory bandwidth of 1.2 TB/s, 3 times that of Nvidia's H100 GPU; at the same time, the chip supports "sparse computing" (Sparse Computing), which can automatically identify and skip invalid calculations to further improve efficiency. These technical advantages make Cerebras' chips a strong competitor to traditional giants like Nvidia and AMD in the AI training field.
2. IPO Background: Surge in AI Computing Power Demand and Market Gap
Cerebras chose to IPO in 2026, driven by the explosive growth in AI computing power demand. According to data from market research firm IDC, the global AI computing power market size will reach $150 billion in 2026, with a compound annual growth rate (CAGR) of 35%. Among them, the demand for AI training computing power is particularly significant, expected to account for 40% of the AI computing power market in 2026, reaching $60 billion.
However, traditional GPU chips have bottlenecks when training large AI models: insufficient computing power density, high energy consumption, and limited memory bandwidth. Cerebras' WSE-2 chip exactly fills this gap, meeting the training needs of large AI models, thus attracting the attention of many tech giants. Before the IPO, Cerebras had established partnerships with companies like OpenAI, Meta, and Google, and its chips were used to train large models such as GPT-4 and LLaMA.
II. Listing Performance: Market Reaction to 150% Surge on First Day
Cerebras' IPO performance far exceeded expectations, mainly due to the following factors:
- AI Hype Continues to Heat Up: In 2026, AI remains the core theme of the US stock market, with leading AI chip stocks like Nvidia and AMD repeatedly hitting new highs, and investors' attention to AI new stocks is extremely high. As a "newcomer" in the AI training chip field, Cerebras naturally became a target for capital chasing.
- Obvious Technical Advantages: The technical parameters of the WSE-2 chip far exceed traditional GPUs, capable of solving key problems in AI training, and the market has high expectations for its future performance.
- Clear Use of Funds: Cerebras plans to use the $1 billion raised for R&D of next-generation chips (WSE-3), expanding production scale, and expanding markets. Among them, R&D investment accounts for 60%, showing the company's emphasis on technological innovation, which meets investors' expectations for tech companies.
- Favored by Institutional Investors: Before the IPO, Cerebras received investments from well-known institutions including SoftBank Vision Fund, Sequoia Capital, and Tiger Global, and the endorsement of these institutions enhanced market confidence.
However, some analysts pointed out that Cerebras' stock price surge was too large, with a callback risk. For example, its price-to-earnings (PE) ratio reached 50 times, far higher than the industry average (about 20 times), mainly because the company has not yet achieved profitability (a net loss of $250 million in 2025). Investors need to pay attention to whether it can convert its technical advantages into actual revenue in the future.
III. Guide to AI New Stock Offerings: How to Seize IPO Opportunities?
For investors, AI new stock offerings need to focus on the following points:
1. Technical Barriers and Competitive Landscape
The AI chip industry has high technical barriers, requiring strong R&D capabilities and patent accumulation. Cerebras' wafer-scale integration technology is its core competitiveness, but giants like Nvidia and AMD are also deploying similar technologies (such as Nvidia's "DPU" chips, AMD's "Instinct MI300"), so it is necessary to pay attention to whether Cerebras can maintain its technological leadership.
In addition, the competitive landscape of the AI chip market is changing: traditional GPU giants (Nvidia, AMD) dominate, but ASIC (Application-Specific Integrated Circuit) and FPGA (Field-Programmable Gate Array) manufacturers (such as Cerebras, Blaize) are rising, and investors need to evaluate the market share potential of these new entrants.
2. Performance and Profitability
Most AI new stocks are in the growth stage and have not yet achieved profitability, so it is necessary to pay attention to their revenue growth rate and gross margin. Cerebras' revenue grew from $50 million in 2023 to $280 million in 2025, with a CAGR of 120%, showing strong growth momentum. However, the gross margin was only 25% (in 2025), lower than Nvidia's 60%, and it is necessary to pay attention to whether it can improve profitability through economies of scale.
3. Use of Funds and Strategic Planning
The use of funds is an important indicator to judge the company's future development direction. Cerebras will use 60% of the raised funds for R&D, showing its emphasis on technological innovation, which is in line with the "technology-driven" characteristics of the AI chip industry. In addition, the company plans to expand overseas markets (such as China, Europe), which will provide impetus for future revenue growth.
4. Market Sentiment and Valuation
The valuation of AI new stocks is often greatly affected by market sentiment. When the AI hype continues, valuations will be high; when market sentiment cools, valuations will callback. Investors need to pay attention to the Federal Reserve's monetary policy (such as interest rate changes), the performance of tech giants (such as Nvidia's financial reports), and policy changes in the AI industry (such as export controls), as these factors will affect the performance of AI new stocks.
IV. Future Outlook: Opportunities and Challenges for AI New Stocks
In 2026, AI new stocks are expected to become a "new hot spot" in the US stock market. According to data from market research firm PitchBook, the number of US AI new stock IPOs in 2026 will reach 15, with a total fundraising of over $10 billion, among which AI chips, AI pharmaceuticals, and AI software are the main fields.
For AI chip new stocks, the opportunity lies in the continuous growth of AI computing power demand, and the challenge lies in technical barriers and competitive pressure. As a representative, Cerebras' IPO performance reflects the market's expectations for AI training chips, but it also reminds investors to rationally assess risks.
In addition, AI new stock offerings need to pay attention to the "industry cycle" and "company fundamentals". For example, when the AI industry is in an upward cycle, new stocks perform well; when the industry enters an adjustment period, new stocks may face callbacks. Therefore, investors need to combine industry trends and company fundamentals to make reasonable investment decisions.
V. Summary
Cerebras Systems' IPO is a "landmark event" among US AI new stocks in 2026, and its 150% surge on the first day reflects the strong market demand for AI training chips. For investors, AI new stock offerings need to focus on technical advantages, performance growth, use of funds, and market sentiment, while rationally assessing risks. As the AI industry continues to develop, AI new stocks are expected to become an important investment theme in the US stock market, but careful selection is needed to avoid blindly following the trend.
