An Interview with MatX CEO Reiner Pope About LLM Chips

Hybrid SRAM + HBM, MoE interconnect, why frontier labs consider AI ASIC startups, and more

An Interview with MatX CEO Reiner Pope About LLM Chips Hybrid SRAM + HBM, MoE interconnect, why frontier labs consider AI ASIC startups, and more Austin Lyons Apr 09, 2026 22 2 Share This interview is with Reiner Pope, co-founder and CEO of MatX. Pope and his co-founder Mike Gunter left Google — Pope from the Brain team, Gunter from the TPU team — one week before ChatGPT launched to build what they believe will be the best chips for LLMs that physics allows. The company has raised ~$600 million to date. In this interview we discuss why Pope left Google to start a chip company, how to overcome

saved by

• Aryan Naik
• Alex K

related reading

• Reiner Pope of MatX on accelerating AI with transformer-optimized chipscheekypint.substack.com
• Interview with Richard Ho, OpenAImorethanmoore.substack.com
• AI Chip Architecturesjepeake.com
• Reiner Pope – The math behind how LLMs are trained and serveddwarkesh.com
• Chip design from the bottom up – Reiner Popeyoutube.com
• How GPT, Claude, and Gemini are actually trained and served – Reiner Popeyoutube.com
• China's AI Chip Landscape: A Complete Map of Who's Building Whatcrossingriver.substack.com
• Are you a memory_man, Part 2 - by Investing with Martinmartinbradstreet.substack.com
• The Inference Shift – Stratechery by Ben Thompsonstratechery.com
• MatX: High-throughput chips for LLMsmatx.com
• Meta harness makes 10 times better Kimi K3 chipluoluo.ai
• MatX AI Chip Startup Raises $500M to Challenge Nvidia | 2026entrepreneurloop.com