AI Summary
5 min readAndrew Feldman, CEO and founder of Cerebras, started the company in 2016 with a simple computer architect’s instinct: when a new workload emerges, you ask whether you can build something faster and whether there is enough of that work to justify the effort. AI was computationally intensive in a way that previous shifts—like mobile processors, which were power-constrained—were not. The incumbent, Nvidia, already had high margins and a long lead. Feldman’s operating principle was that being a little better or a little cheaper was not an available strategy against a giant that could cut prices or bundle products. The only viable path was radical innovation: something ten, a hundred, or five hundred times faster, with a value proposition that held even if the competitor gave its product away.
Radical innovation means doing everything yourself
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of Uncapped with Jack Altman
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 (01:07) **Origin Story: Why Cerebras Started in 2016** - Andrew explains the computer architect's thought process: spotting a new, computationally intense workload (AI) and asking "Can I build a faster chip?" and "Is there enough work to justify it?"
- 2 (04:30) **The "Attack Goliath" Strategy: Why Radical Innovation is the Only Path** - Andrew lays out the core philosophy: being "a little bit better" or "a little bit cheaper" than Nvidia is not a viable strategy because they have high margins and can bundle or cut prices.
- 3 (07:20) **The 100x Rule: Why a 10x Improvement Isn't Enough** - Eric explains the math behind why new chip companies need a massive advantage to survive the long development timeline.
- 4 (09:55) **The Near-Death Experience: 18 Months of "Still Can't Make It"** - Andrew describes the brutal period where the company was spending $8 million a month and failing to make wafer-scale technology work, with no progress to report to the board.
- 5 (13:08) **The Next Wafer-Scale Innovation: Memory and I/O** - Andrew outlines the three fundamental elements of a computer (calculation, memory, I/O) and where Cerebras is pushing the R&D envelope next.
- 6 (16:25) **The Chip Supply Chain Explained: From Sand to a $22 Chip** - Andrew provides a clear, high-level explanation of the chip manufacturing process, from TSMC's fabs to ASML's lithography machines.
- 7 (21:53) **The Fab Bottleneck: Why It's Hard to Build More** - Andrew and Eric discuss why TSMC's capacity is a huge constraint and the strategic vulnerability this creates for the US.
+ Full timestamped outline available in the app
Guests on this episode
Show Notes
Andrew Feldman is the co-founder and CEO of Cerebras Systems, the AI chip company he founded in 2016 around a single radical insight: that winning in compute requires not incremental improvement but a fundamentally different architecture. Cerebras is the creator of the world's largest chip, the Wafer Scale Engine, and counts the US government, sovereign cloud providers, and OpenAI among its customers.
Alongside Eric Vishria from Benchmark, we discussed why Andrew believes that if you are going to attack Goliath, being 10% or even twice as good is not an available strategy and you have to aim for 100x or 500x better. Andrew walked through Cerebras's near-death experience: 18 months of board meetings where the only thing to report was "still can't make it," spending $8 million a month, and what kept the team going. He explained how we go from sand to a ChatGPT answer and why the US semiconductor supply chain is in a precarious position. Andrew shared what most people get wrong about what makes Nvidia great (it’s not CUDA), and why he thinks of himself as a professional David in an ongoing battle with Goliath.
Timestamps:
(0:00) Intro
(1:07) Why Andrew started Cerebras in 2016
(2:54) Eric on why he invested despite having no chip experience
(4:10) Attacking Goliath
(9:44) Near-death experiences and the Valley of Death
(10:44) 18 months of "still can't make it"
(12:16) Solving a 75-year-old compute problem
(13:26) What comes after Wafer Scale
(16:19) The chip supply chain explained
(22:30) Why the US punted a strategic industry
(25:51) How to be a good hardware board member
(27:50) Hardware vs. software investing
(29:05) The pivot from training to inference
(27:00) Specialization vs. flexibility
(35:22) Young product leaders and seasoned hardware engineers
(39:14) External relationships and TSMC
(42:20) The AI infrastructure buildout
(44:12) The data center supply chain
(53:14) Speed creates markets
(54:10) Disaggregation with AMD and AWS
(55:24) What actually makes Nvidia great
(57:30) Near-death experiences and the DNA of a Goliath fighter
(59:58) Andrew's childhood next to William Shockley
Links:
https://x.com/ericvishria
https://x.com/jaltma
https://uncappedpod.com/
More from this podcast
Uncapped with Jack Altman →