Remember when we told you Cerebras was approaching its IPO at a valuation of $26.6 billion? That was three weeks ago. The company has since decided that number was adorable, priced its shares at $185 — well above the already-raised range of $150–$160 — and watched them close their first day at $311.07 after touching an intraday high of $385.
The $5.5 billion raise is 2026’s largest IPO. The stock surged 68% on day one. The market cap settled around $70 billion on outstanding shares, or roughly $86 billion fully diluted. Demand exceeded available shares by more than 20 times.
Allow us to repeat that: a company whose primary product is a chip the size of a dinner plate just debuted at a valuation larger than most European nations’ GDP.
🤚 The Open-Palm Offering
Cerebras Systems (CBRS) listed on the Nasdaq on May 14, completing the longest-awaited AI hardware IPO since investors started hallucinating about silicon alternatives to Nvidia. The company initially planned to raise around $4 billion but bumped the offering to 30 million shares at $185 each after order books filled 20x over.
Key numbers:
- IPO price: $185 per share
- First-day close: $311.07 (+68%)
- Intraday high: $385 (+108%)
- Total raised: $5.55 billion
- Market cap: ~$70B (outstanding), ~$86B (fully diluted)
For context, when Cerebras filed its S-1 last year, its last private valuation was $4.25 billion. It is now worth roughly twenty times that. This is what happens when you spend eight years building a wafer-scale chip and then the entire planet decides it needs one yesterday.
👐 The Two-Handed Market Frenzy
The obvious question: is a company that competes with Nvidia — a $3.5 trillion colossus with a near-monopoly on AI training hardware — actually worth $86 billion before it has posted a profitable year?
The market’s answer, delivered with the subtlety of a man buying a yacht on margin, was: yes, obviously, have you seen the TAM slides?
Cerebras’s wafer-scale engine — the WSE-3 — is a single chip the size of an iPad that contains 4 trillion transistors. It solves the memory-bandwidth problem that makes large model training on GPU clusters feel like pouring the ocean through a garden hose. Hyperscalers have noticed. AI labs have noticed. The Department of Energy has noticed. And now, apparently, every retail investor with a Robinhood account has also noticed.
The 20x oversubscription suggests institutional money is desperately seeking the “second source” narrative — the idea that customers who currently write Nvidia a blank check every quarter would very much like a second option, even if that option requires rethinking their entire rack architecture.
🌿 The Gentle Awakening
There is something philosophically wonderful about a company whose entire thesis is “what if we just made the chip bigger” achieving an $86 billion valuation. In an industry obsessed with architectural cleverness, novel interconnects, and exotic materials, Cerebras essentially walked into the room and said: “What if we didn’t cut the wafer?”
And the market said: “Here is $86 billion. Please continue not cutting the wafer.”
This is either a vindication of first-principles engineering or the greatest proof yet that investor FOMO is the most powerful force in the known universe — more powerful than gravity, more persistent than inflation, and entirely immune to price-to-earnings ratios.
👑 The Gold-Leaf Reckoning
What Cerebras’s debut really tells us is that the AI hardware market has graduated from “Nvidia and friends” to “Nvidia and challengers with real revenue.” The IPO window — frozen since 2022 — didn’t just open; it was blown off its hinges by a company that makes chips requiring their own postal code.
For Nvidia, this is a compliment wrapped in a competitive threat. For AMD, Intel, and every custom-silicon startup in a WeWork, it’s a signal: the public markets are hungry for alternatives, and they will pay irrational premiums for the privilege of owning one.
For the rest of us, it’s a reminder that in the AI gold rush, the companies making the shovels are now worth more than most of the mines. Cerebras just proved you can IPO a chip company at a valuation that would make a mid-cap pharma weep — and the market will still ask for seconds.
“They said ‘make the chip bigger’ and the market said ‘make the valuation bigger’ and honestly, who are we to argue with $86 billion in fully diluted conviction?” — The Slap of Wisdom Trading Floor, currently trying to fit CBRS shares into a portfolio that was already 60% Nvidia