Cerebras US$5.55B IPO fixed

The #1 Number Cerebras’ CFO Fixed Before Their $5.55B IPO That Most Investors Don’t Know Exists

One detail from the Cerebras S-1 that nobody’s talking about: Bob Komin — their CFO — joined in March 2024. The IPO closed May 2026. That’s 26 months of runway before the bell rang. Bob wasn’t hired to execute the IPO. He was hired to build the infrastructure that made the IPO survivable.

=====

Cerebras just closed its IPO up 68% on Day 1.

I kept thinking about one thing: their data room.

$5.55 billion raised.

First-day close at $311 — up from a $185 price that was already above range. The Cerebras IPO just became the largest US tech offering since Uber’s 2019 debut.

The AI chip story is compelling. But as a CFO, I wasn’t watching the ticker — I was thinking about what their finance team had to build to get there.

Because here’s what most founders don’t see from the outside: An IPO isn’t a fundraise. It’s a finance audit with a public audience.

And the companies that stumble aren’t the ones with bad products. They’re the ones with beautiful revenue and brittle reporting infrastructure underneath it.

I’ve been in the room when a founder’s fundraise quietly died. Not because of their product. Because of one question from a Series B investor they couldn’t answer cleanly.

The question wasn’t complex.

“Can you walk me through your revenue recognition policy — and show me where that decision lives in your system?”

The business owner — genuinely a great operator, real product-market fit, strong growth — paused. Then said something that ended the conversation: “I’d need to get back to you on the exact mechanics.”

That pause cost them the cheque.

Here’s the thing nobody tells you about raising growth-stage capital: investors aren’t buying your growth rate. They’re stress-testing the infrastructure beneath it. Because they’ve seen enough companies where the headline number was real but the system behind it couldn’t survive diligence.

3 Lies SME owners tell themselves before an IPO:

1 “Our numbers are clean because our accountant signs off on them.”

An accountant signing off on your year-end accounts is not the same thing as having IPO-grade financial reporting. Not even close.

IPO-grade reporting means:

• Monthly financials that are consistent, comparable, and auditable going back three years
• Revenue recognised under the correct accounting standard (IFRS 15 or ASC 606) — not the standard that makes your numbers look best
• Related-party transactions disclosed, documented, and defensible
• A chart of accounts your auditor and your underwriter can read independently

What most growth-stage companies have: annual accounts that are technically correct, prepared retrospectively, in a format that works for your local reporting as a private company but would require significant restatement to meet listing standards.

That restatement process — under SEC, SGX, HKeX, ASX, or LSE scrutiny — takes 12 to 18 months, costs more than most business owners budget for, and is the single biggest timeline killer in IPO preparation.

The fix isn’t complicated. But it requires someone in the room who’s been through it.

This is where I spend most of my time as a CFO-partner with SME founders. Not on the pitch deck. On the infrastructure that makes the pitch defensible.

2 “We’ll hire a proper CFO when we’re closer to IPO.”

This is the most expensive mistake in the playbook.

The CFO you need for an IPO is not the CFO you can hire six months out. Because a CFO hired six months before your target listing date inherits the infrastructure problem, doesn’t have enough context to fix it quickly, and gets blamed for delays they didn’t cause.

3 “Our growth rate will paper over any reporting gaps.”

This was true in 2021. It is not true in 2026.

Post-rate-cycle, institutional investors — and underwriters — have viscerally recalibrated their risk tolerance for companies with strong growth but weak governance infrastructure. The companies that received outsized multiples in 2021 without clean reporting did so in a market where capital was cheap and competition for deals was fierce.

That market is gone.

What’s replaced it: investors who’ve spent three years marking down positions in companies where the growth story was real but the infrastructure wasn’t. They are now — structurally, institutionally — more skeptical of exactly the signals that used to get waved through.

The phrase I hear most in investor conversations right now: “We want to see institutional-quality reporting before we get to pricing discussions.”**

Your growth rate (if any after accounting for the related party transactions) opens the door. Your finance infrastructure keeps you in the room.

=====

Cerebras had 108% intraday pop energy because investors trusted the story. That trust starts in the financial data infrastructure — not on the roadshow stage.

Bob Komin has served as Cerebras’ CFO since March 2024 — and the IPO closed May 2026. He previously led Sunrun’s IPO as CFO, and took Flurry through its Yahoo acquisition and Tellme Networks through its Microsoft acquisition. A deliberate, experienced hire — not a last-minute one.

The IPO window is open right now. OpenAI, Anthropic, and a dozen growth-stage AI companies are queuing up behind Cerebras.

If you’re planning an IPO or a significant capital raise in the next 24 months — and you don’t have someone in your corner who has lived that process — what’s your plan for the gaps you don’t know you have?

Because those are the ones that matter most.