Author: AI New Era
Insane! A chip the size of a dinner plate directly enabled the largest tech IPO of 2026. On its first day, it surged 108%, making Altman nearly tenfold richer and launching the first wave of a $3 trillion IPO tsunami in the ASI era.
In 2026, the largest U.S. tech IPO was born!
Today, AI chip company Cerebras officially listed on Nasdaq, surging 108% on its first day of trading.
The IPO was priced at $185 per share, opened directly at $385, and closed at $311, with a valuation that once reached $100 billion.
Cerebras sold 30 million shares in a single offering, raising $5.55 billion.
This is one of the largest U.S. tech IPOs since Uber's 2019 listing, even outpacing the momentum of Snowflake back then.
At the Nasdaq trading floor, Andrew Feldman, founder and CEO of Cerebras, rang the opening bell.
Today, a large number of VC investors have become millionaires overnight; Altman holds 89,000 shares, multiplying his stake tenfold to approximately $30 million.
Cerebra made a stunning debut, successfully igniting the first AI IPO of 2026; next in line are OpenAI, SpaceX, and Anthropic.
The big surge in AI IPOs is about to begin.
2026's largest AI IPO, surging 108% on its first day
To be honest, a year ago, no one believed this day would come.
In 2015, semiconductor industry veteran Andrew Feldman co-founded Cerebra with seasoned chip industry experts.
When Feldman started selling chips in 2019, he himself said, "Nobody cares, the market isn't ready."
But saying that Cerebras’s path to listing is a suspense drama is no exaggeration.
In September 2024, Cerebras submitted its IPO application for the first time.
As a result, it came under the scrutiny of the U.S. Committee on Foreign Investment (CFIUS)—
The reason is that its major client and investor, G42, is based in Abu Dhabi and at the time contributed 87% of Cerebras's revenue.
The national security review has been delayed repeatedly, investor confidence has collapsed, and the IPO plan has been forced to be put on hold.
A year later, the plot has completely reversed.
AI models have finally become "smart and useful," and Cerebras' business has surged directly:
Revenue of $510 million in 2025, a 76% year-over-year increase. More importantly, the company turned a loss of $482 million into a profit of $238 million.
In just one year, a turnaround from massive losses to huge profits has left investors completely unable to sit still.
The customer list has shifted from being dominated by G42 to including multiple major players such as OpenAI, AWS, G42, and MBZUAI.
More importantly, two major new clients have emerged—
OpenAI has signed a multi-year contract worth over $20 billion to secure 750 megawatts of Cerebras computing power, to be deployed in phases starting in 2026.
AWS has announced the deployment of Cerebras CS-3 chips in its own data centers, making them available to developers through Amazon Bedrock.
In a recent interview, Feldman revealed that during the roadshow, the three core ideas he most needed to convince investors of were:
First, inference demand will increase by a factor of one million.
Second, acquiring computing power is not limited to GPUs;
Third, CUDA's moat has been overstated.
This time, Cerebras successfully kicked off the AI IPO campaign with an impressive debut.
In the ASI era, overnight wealth and OpenAI also become "big winners"
Cerebras' IPO created a textbook-perfect VC exit bonanza.
From the seed round to the listing, these institutions waited a full decade.
Foundation Capital: $37 million → $2.8 billion, 76x return
Foundation Capital invested approximately $37 million, holding a 7% stake, which, at the IPO price of $185, is valued at $2.8 billion, representing a 76x return.
At the first-day closing price of $311, approximately $4.8 billion.
Benchmark: 268 million → 3.3 billion, 12x return
Benchmark also entered in 2016 during the Series A round, led by General Partner Eric Vishria.
It invested a total of $268 million, holding an 8.1% stake; at the IPO price, this equated to a $3.2 billion valuation and a 12x return. Based on the closing price, it surged directly to $5.5 billion.
Eclipse Capital: $146.5 million → $2.5 billion, 17x return
Eclipse Ventures invested $146.5 million, holding a 6.2% stake, at an IPO valuation of $2.5 billion, representing a 17x multiple.
The story of Eclipse is more legendary, with the investment decision made by Pierre Lamond, a 95-year-old veteran of Fairchild Semiconductor.
Lamond invested in Feldman’s previous company, SeaMicro, while at Khosla Ventures, and later joined Eclipse.
A year after Cerebras was founded in 2016, he decided to invest again. This was equivalent to the godfather of Silicon Valley chips personally endorsing Cerebras.
In 2017, he said, "Feldman is one of the few entrepreneurs I would invest in twice."
Ultraman 10x "passive income," invest at just $0.01
Among all VC returns, the most outrageous is OpenAI!
First, let’s talk about Orman personally.
According to court documents that were made public, Altman personally invested in Cerebras as early as February 2017, long before ChatGPT existed.
By the end of 2025, he held 89,373 shares, worth approximately $3.2 million at that time.
On its first day of listing, the value of these shares surged to approximately $30 million.
These funds quietly grew tenfold.
Previously, OpenAI once considered directly acquiring Cerebras.
Although the acquisition fell through, both parties reached a "secret agreement" on Christmas Eve—
Over the next few years, OpenAI could acquire up to an 11% equity stake by purchasing billions of dollars worth of inference computing resources from Cerebras.
OpenAI paid less than one cent per share for these stakes.
With the IPO price set at $185, the value of OpenAI's holdings surged instantly. Currently estimated, OpenAI has an unrealized gain of approximately $1.8 billion.
Notably, OpenAI co-founder Greg Brockman also held approximately 77,000 shares of Cerebras in his personal capacity.
Based on the closing price of $311, Brockman's personal holdings are worth approximately $24 million.
Technical advantage: a chip the size of a dinner plate
It is well known that Cerebras' core weapon is the WSE-3 (Wafer Scale Engine 3).
A single chip occupying an entire 300mm wafer, with an area of 46,225 square millimeters—about the size of a dinner plate.
A typical chip is about 800 square millimeters, while the WSE-3 is nearly 60 times larger.
In terms of specifications, WSE-3 achieves a decisive advantage—
4 trillion transistors (19 times that of the NVIDIA B200)
900,000 AI-optimized cores
125 petaflops of AI computing power (28 times that of the B200)
5nm process, manufactured by TSMC
While others are tackling compute power issues with multi-GPU setups, Cerebras turns an entire wafer into a single massive processor.
Behind this brutal aesthetics lies a key insight: reasoning is the future battleground of AI.
Training a large model may only need to be done once.
But reasoning—having the model answer every question and execute every Agent task—is a continuous, unceasing consumption of computational power.
Cerebras claims their inference speed is 10 to 20 times faster than NVIDIA GPU clusters.
AI IPO tsunami of $3 trillion is poised to hit
Keep in mind, the Cerebras IPO is just an appetizer.
The real highlight is yet to come—
SpaceX (including xAI): Roadshow as early as June, valuation target of $1.75 trillion, plan to allocate 30% to retail investors—an unprecedented move in IPO history.
If successful, it will be the largest IPO in human history.
OpenAI: Targeting Q4 2026 for IPO, aiming for a $1 trillion valuation.
Just completed the largest single private placement in history, raising $122 billion. But with $13.1 billion in revenue in 2025 and a loss of $14 billion, profitability is not expected until 2030.
Anthropic: Planning to go public in October, has already raised $30 billion in funding, with a valuation of $900 billion; IPO proceeds may exceed $60 billion.
According to the latest data, its annualized revenue has surged from $9 billion at the end of 2025 to over $44 billion.
The combined valuation of the three exceeds $3 trillion.
Fortune wrote a bold statement: "SpaceX, OpenAI, and Anthropic could either reignite the IPO market or drain it dry."
|
|
If three companies list within six months, their combined fundraising could exceed $240 billion, surpassing the total U.S. IPO fundraising in most years in recent times.
The Final Showdown: When Capital Begins Betting on ASI
But if you think Wall Street is merely betting on a chip company’s financial statements, you’re being naive.
Looking at the timeline more broadly, Cerebras’ story reveals a deeper logic: capital is betting on ASI in advance.
Back in 2017, Greg Brockman wrote in an email to the OpenAI team:
Exclusive access to Cerebras' hardware will give OpenAI a decisive computational advantage over Google.
At that time, ChatGPT did not yet exist, and LLMs were still concepts in academic papers, but OpenAI’s core team already knew one thing clearly:
The path to AGI and ultimately ASI is ultimately a computing power arms race. Today in 2026, this assessment is being wildly validated.
What was truly at stake in this IPO was never just a company's price-to-earnings ratio.
It is betting on the next form of intelligence that humanity is building—and the computing power required for this intelligence far exceeds what anyone can imagine today.
When capital begins valuing a 10-year-old chip company at $66 billion, when OpenAI is willing to offer $20 billion just to secure 1/40th of the compute gap, and when SpaceX, OpenAI, and Anthropic are lining up for trillion-dollar IPOs—
This is humanity laying the groundwork for the arrival of superintelligence.
Whoever lays the foundation first secures their ticket to the ASI era. Whoever hesitates will forever remain only an observer.