May 21, 2026 · 30m · no-priors
The Story Behind Cerebras’ $63 Billion IPO with Founder and CEO Andrew Feldman
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this interview on the No Priors podcast, Cerebras co-founder and CEO Andrew Feldman details the company's journey from pioneering radical wafer-scale AI chips amid intense skepticism to achieving a $63 billion IPO and securing historic hyperscale partnerships. He articulates how ultra-fast inference speed will fundamentally reorganize global enterprise workflows and unlock new technological paradigms.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 19.6% of the talking time here. How this is scored →
The hosts as informed peer 5.3 Guest teaching 4.0 Guest disagreement 1.3 The hosts pushing back 1.6
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 16:45 Embracing the David vs Goliath chip on the shoulder
Andrew passionately rejects the notion of building tech just for financial gain, framing Cerebras's entire ethos around outsmarting Nvidia's muscle with contrarian engineering.
Hardest push from the hosts ▶ 18:27 Elad challenges clean hypothesis-testing frameworks for shutdown decisionsElad pushes back against Andrew's tidy formulation for quitting, pointing out that in practice founders continuously shift goalposts by testing 'just one more thing' sequentially.
Biggest teaching moment ▶ 7:14 The 70-year history of failed wafer-scale chips and Gene AmdahlAndrew educates the room on the computing industry's failed attempts to achieve wafer-scale integration, highlighting computing pioneer Gene Amdahl's failures and Cerebras's 2-year struggle.
The host holds their own ▶ 5:11 Elad maps out the mid-2010s machine learning landscapeElad showcases deep institutional knowledge of the early deep learning ecosystem, listing CNNs, RNNs, and nascent GANs when Cerebras was first being founded.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| Cerebras Overview, IPO Milestone, and Explosive Demand | 4 | 3 | 1 | 1 | Elad opens by celebrating Cerebras's recent $60B+ IPO and asks targeted questions about their 15-20x inference speed advantage and explosive market shift over the past 18 months. Andrew cooperatively details their recent massive contract wins with OpenAI and AWS. | |
| Betting on Radical Wafer-Scale Architecture | 5 | 4 | 2 | 3 | Sarah challenges the original perception of Cerebras, noting critics called wafer-scale a weird architecture that had it wrong. Andrew embraces the contrarian framing, explaining why radical speed gains require totally non-derivative hardware. | |
| Early Conviction and Overcoming Technical Failure | 6 | 4 | 1 | 1 | Elad demonstrates domain familiarity with 2016-era ML architectures like CNNs and GANs while inquiring about conviction during dark periods. Andrew candidly recalls spending $8M a month while facing repeated silicon failures between 2017 and 2019. | |
| Surviving the Market Chasm with Strategic Partners | 5 | 5 | 1 | 1 | Sarah and Andrew discuss bridging the adoption chasm through supercomputing centers and a pivotal $1B contract with G42. Andrew explains the physical realities of scaling hardware QA labs and hardware production constraints. | |
| Compiler Complexity and Internal AI Engineering Workflows | 5 | 4 | 1 | 1 | Sarah brings up compiler complexity and internal workflow automation, recalling early pitch meetings. Andrew details Cerebras's internal shift from $1k to $30k per engineer on LLM agent tokens. | |
| Preserving Fearless Culture, Psychology, and Quitting Discipline | 6 | 4 | 2 | 3 | Elad pushes back on neat formulas for shutting down ventures, arguing founders fall into sequential testing traps. Andrew elaborates on fighting Goliath (Nvidia) and having trusted peers hold founders accountable to avoid the boiling frog syndrome. | |
| The IPO Strategy and the Historic OpenAI Agreement | 6 | 4 | 1 | 1 | Elad and Andrew discuss the evolution of IPO horizons and standard 4-year option vesting origins. Andrew explains why going public provided unique AI pure-play status and reviews closing the OpenAI contract in just four weeks. |
Statements from this episode (15)
Feldman: Cerebras is 15 to 20 times faster than GPUs at inference
“And right now we're the fastest at inference, not by a little bit, but by a lot. 1518, 20 X faster than GPUs.”
Feldman: Cerebras signed an agreement to deploy in AWS data centers
“And then in March, we signed an agreement with AWS, where we will be deployed in their data centers going forward”
Feldman: The market for slow AI inference will become zero
“How big is the market for slow search? It's zero. How big is the market for dial-up internet? It's zero. That's how big the market for slow inference will be.”
Feldman: Computer pioneer Gene Amdahl failed miserably at wafer-scale compute
“They've been efforts across the entire seventy-year history of the compute industry to build a wafer scale product. In fact, Gene Amdahl, sort of one of the fathers of our field, one of the guys on Mount Rushmore of compute, failed miserably to do it.”
Feldman: Cerebras burned $8M monthly for two years before chips worked
“We had a period between about 2017, middle of 2017 and middle of 2019 where we couldn't build it. We were spending about eight million a month.”
Feldman: Cerebras will sell tens of thousands of third-generation systems
“It was like, you know, the first gen we might have sold a dozen. The second gen we probably sold 300, and now we're still going to sell tens of thousands in the third gen.”
Feldman: New chip architectures must start in supercomputing where speed trumps software maturity
“I think there's a path that has been laid down by new computer architectures, and often you begin in the supercomputer world, because those guys love speed, and they don't care if your software is immature, and so we sort of ran the table there.”
Feldman: UAE sovereign AI firm G42 placed a $1B order with Cerebras
“And we won a sovereign, a G-forty-two and they became a strategic partner and close friends and they placed a billion dollar order on us,”
Feldman: Building a compiler realistically takes about 10 years
“It takes about 10 years. It takes a long time to build a compiler. Is an extraordinarily difficult piece of software.”
Feldman: Cerebras spends $25k to $30k per engineer on AI tokens
“I would say that, that, you know, eight months ago, we weren't spending a thousand dollars in engineer on tokens, and we're probably at 25 or 30,000 right now, and it's ripping.”
Feldman: Companies scaling past 1,000 employees dangerously abandon fearless risk-taking
“I think one of the malaise of companies as they get to a thousand to 2003 thousand people is they stop taking the type of risks. That they were taking before, right? You move from being a fearless engineering culture to sort of being, what can we get in in the…”
Feldman: Cerebras is the only pure-play public AI company
“We would be the first and only, for a period of time, AI pure play. We are the only company that, that you can, a hundred percent of the revenue, this exact market. There's no gaming, there's no graphics, there's no PC, that, this is it.”
Feldman: I refuse to build or sell consumer products
“I have a rule that if my mother buys it or uses it, I don't want to make it or sell it. Cause I really want super smart customers who are doing really interesting things with our stuff.”
Feldman: Cerebras signed a $20B+ OpenAI deal in 4.5 weeks
“For a, 20 plus billion dollar deal to do it in four and a half weeks was exceptional.”
Feldman: Chinese open-source AI techniques force closed labs to innovate
“The open source community has sort of kept the interest alive and kept the flame going. And I think that, that the and pushed the closed source guys. I think the sort of techniques that we saw by some of the Chinese Like, whoa, we gotta stay ahead of that, rig…”