Cerebras Systems (CBRS), a rising star in the artificial intelligence chip sector, saw its stock price surge 68% on its first day of trading last week, drawing significant market attention. The logic behind the market’s enthusiasm is clear: Cerebras offers a product in high demand—high-performance AI chips. The company even claims that its chip performance surpasses systems based on graphics processing units (GPUs), which are the flagship products of Nvidia (NVDA), the leader in the AI chip market.
Leveraging its market dominance, Nvidia has delivered substantial returns for investors in recent years, with its stock price climbing 1,400% over the past five years. Behind this performance lies steady profit growth: AI companies scrambling to purchase computing power from Nvidia have driven its revenue into the tens of billions of dollars. In its most recent full fiscal year, this AI giant reported revenue exceeding $215 billion, a 65% increase year-over-year.
Now that Cerebras has just become a publicly traded company, people are asking: Can this young AI firm become the next Nvidia?
Cerebras builds chips that are 58 times larger than Nvidia’s B200 chip, and this immense size delivers extremely high speeds. The company states that its Wafer-Scale Engine (WSE) is 15 times faster in inference than GPU-based systems, and in some cases, up to 1,000 times faster. Cerebras’s WSE-3 contains 4 trillion transistors, whereas a package consisting of two of Nvidia’s GPUs has only 208 billion transistors.
Customers can access the WSE by ordering platforms for their data centers or through the Cerebras Cloud and third-party cloud services. As customers flock to the product, Cerebras’s revenue has grown dramatically: from 2022 to last year, revenue surged approximately 2,000% to reach $510 million.
Investors are clearly optimistic about Cerebras’s progress in the AI market. The company saw its stock jump on its debut on May 14 and raised $5.5 billion, marking the largest IPO of the year. However, in subsequent trading sessions, the stock gave back some of those gains.
Cerebras operates in the same space as market leader Nvidia, both striving to make customers’ AI tasks faster and more efficient. However, the two companies have vastly different histories. Nvidia was founded over 30 years ago and has spent many years honing its GPU technology and market presence. GPUs initially served the video game market, but after recognizing the broader potential of its chips, Nvidia built the CUDA parallel computing platform, paving the way for its subsequent expansion. Later, Nvidia saw the AI opportunity and began designing GPUs specifically for this market.
In contrast, the younger Cerebras, founded in 2015, has focused from the very beginning on providing solutions for the immense computing power demands brought by AI. Today, Nvidia continues to serve the gaming market and has expanded into AI, offering solutions for specific industries ranging from robotics to healthcare. Nvidia’s largest customers are tech giants like Microsoft and Amazon, but it also serves smaller enterprises.
Cerebras’s revenue is growing rapidly, but the problem lies in its revenue concentration among a small number of clients, creating vulnerability. To become the next Nvidia, Cerebras must significantly increase its customer base, incorporating major U.S. tech giants and other international companies. At the same time, the company must genuinely prove its advantages in AI over the long term, reflecting this in its financial reports through revenue growth and eventual profitability. However, building an empire is no easy task for a newcomer like Cerebras—after all, Nvidia’s rise was the result of decades of research, development, and growth.
Looking at stock performance, history suggests that newly public stocks do not necessarily outperform the broader market in their initial years. Research by Jay Ritter, a finance professor at the University of Florida, shows that excluding first-day gains, IPO stocks underperform comparable-sized companies by 3.6% annually in the five years after going public. If historical patterns hold, Cerebras may find it difficult to replicate Nvidia’s trajectory of gains.
If Cerebras can consolidate its strengths and achieve expansion in its customer base and industry coverage, it still has the potential to grow into a new industry leader in the future.
