Cerebras CEO Andrew Feldman x Alex Wissner-Gross: Wafer-Scale Engines, Fabs as Pyramids | ASI Pill EP256

<p>Special drop: Cerebras CEO Andrew Feldman, fresh off the $95 billion IPO, in conversation with Alex Wissner-Gross. Alex asks the questions, Andrew tells the story.</p><p><br></p><p>A wafer-scale engine 58 times larger than any chip ever built. SRAM stuffed to the gills on a chip the size of a dinner plate. The contrarian bets that took years of perseverance, $8 million a month for 18 months with no solution, then breakthrough. 15-20x faster than the GPU on inference. The OpenAI deal north of $20 billion. The AWS term sheet.</p><p><br></p><p>Then the bigger questions: how do you cut a 10-trillion-parameter model across chips? Why is Cerebras the right platform for chips in orbit, where fault tolerance is everything? Why do fabs take five years and $50 billion, with TSMC the greatest manufacturing company on Earth? Why do Samsung and Intel keep failing at the same node? Why are NVIDIA and AMD great at GPUs and terrible at everything else?</p><p><br></p><p>Alex pushes on Sam Altman as the counterweight to Elon and SpaceX AI's Dyson Swarm. Andrew on luck, hard work, and why you don't need 'the highest paid talent' to win. The pressure test on the soul of being a startup CEO.</p><p><br></p><p>Built from the Moonshots Podcast EP256 (Peter Diamandis, Salim Ismail, Dave Blundin), where Andrew Feldman joined Alex Wissner-Gross. Timestamps below.</p><p><br></p><p>#AI #ASI #ASIPill #Cerebras #AndrewFeldman #AlexWissnerGross #WaferScale #SemiconductorIPO #Inference #SpaceCompute #Moonshots #Hyperstition</p><p><br></p><p>TIMESTAMPS:</p><p>0:00 Welcome and setup</p><p>0:17 [Parzival] Strap in. The Cerebras CEO just rang the bell for a 95 billi</p><p>0:23 Andrew on the $95B IPO bell</p><p>0:48 [Parzival] And what does Andrew make of Karpathy joining Anthropic, and</p><p>0:54 Karpathy and frontier-lab drift</p><p>2:10 [Parzival] And what does the Cerebras founder think of the Elon and Sam</p><p>2:14 Elon vs Sam, both building cool</p><p>3:19 Cerebras mojo, exceptional people</p><p>3:35 [Parzival] On the ASI Pill, we want the origin story. How do you actual</p><p>3:43 Founding bet: chip the size of a dinner plate</p><p>8:30 [Parzival] And which of all those bets does Andrew think were dead wron</p><p>8:33 Got many bets wrong, inference bet right</p><p>9:51 Alex Q: SRAM at trillion-parameter scale</p><p>10:25 Cleaving 10T-parameter models across chips</p><p>14:06 [Parzival] And what about Elon's Tera Fab? How long does that actually </p><p>14:10 Elon's fab bet: 15-20 year project</p><p>16:15 [Parzival] Then what is so hard about building a fab, really?</p><p>16:18 Fabs are pyramids</p><p>19:31 [Parzival] And which fab does Cerebras actually use, and which one woul</p><p>19:35 Three-nanometer at TSMC, never Intel</p><p>21:22 [Parzival] What is the thing Cerebras quietly solved seven years before</p><p>21:26 Solved CoaS seven years before NVIDIA hit it</p><p>25:18 Alex Q: WSE 8 and the 10-year future</p><p>25:51 Infrastructure builders don't pick the apps</p><p>27:52 Alex: deliberately deferring to frontier labs</p><p>28:09 Sparse linear algebra is the bet</p><p>28:21 [Parzival] And what is the killer app for the wafer-scale engine in spa</p><p>28:25 Why Cerebras has an edge in space</p><p>29:53 Alex: fault tolerance in ionizing radiation</p><p>30:19 Production chips in space, a decade out</p><p>31:29 [Parzival] Which professions actually fall first to language models?</p><p>31:32 Lawyers and accountants vs language models</p><p>32:23 Alex teases Andrew on obscure knowledge</p><p>32:33 Why Intel and AMD failed at cellphone chips</p><p>33:57 Alex: massively transformative purpose</p><p>34:02 What made Intel say no to Apple</p><p>34:15 Intel chasing margins, sold its ARM division</p><p>34:23 Luck is not equally distributed</p><p>35:01 [Parzival] And why does the team with the most money lose, again and ag</p><p>35:04 Why the team with the most money doesn't win</p><p>36:24 Alex: Sam Altman as counterweight to Elon</p><p>37:12 You underestimate Sam at your cost</p><p>38:58 Alex: fabs in space don't need heavy launch</p><p>39:58 Building a fab on land is hard enough</p><p>40:25 [Parzival] What is China actually winning at, when it comes to AI infra</p><p>40:29 China's power infrastructure advantage</p><p>41:17 Alex: China siphoning American tokens</p><p>41:43 Being CEO is enough</p><p>41:49 Signoff</p>