Our mission

Make the world’s best AI models run as efficiently as allowed by physics, bringing the world years ahead in AI quality and availability.

This enables a happier and more prosperous world, where for example everyone can have a huge AI staff of specialist MDs, tutors, coaches, advisors, and assistants.

Press

Read MatX’s profile on Bloomberg.

Our team

• Reiner Pope (cofounder and CEO) has been building high performance software and hardware for over a decade, across the stack from ML chips to distributed system infrastructure to LLMs.
  • As Efficiency Lead for Google PaLM, he designed and implemented the world’s fastest LLM inference software, and he helped conceive Google’s TPU v5e and optimize it for LLMs.
  • As an Architect for one of Google’s ML chips, he helped design Google’s then-fastest ML chip.
  • As Compiler Lead for one of Google’s ML chips, he has experience solving one of the hardest problems for ML chips (building the software environment), and he knows how to design chips to make this task dramatically easier.
  • Reiner has 11 patents.
• Mike Gunter (cofounder and CTO) is a multi-time founder and has 28 years of experience in hardware architecture, including 12 years of experience in machine learning.
  • As Chief Architect for one of Google’s ML chips, he designed Google’s then-fastest ML chip and invented techniques to improve latency by orders of magnitude.
  • As an Architect for Google’s TPUs, he developed optimizations for the core computation units.
  • Mike has designed and implemented 11 chips for applications across processors, wireless communications, graphics, and ML, including Google’s first hardware accelerator.
  • Mike has 23 patents.
• Avinash Mani (Chief Development Officer, Silicon) has over 25 years of experience building products and world-class engineering teams in silicon and software.
  • He has taken over 12 chip products from concept to high-volume customer deployment.
  • Avinash has built and led multiple teams of 100+ engineers at Amazon, Innovium, and Broadcom, across cloud infrastructure, ML, consumer devices generating billions of dollars in value.
  • He was on the founding team at Innovium and helped build the company from inception to a $1.1B exit.
  • Avinash has over 10 patents.
• Everyone on the MatX team is exceptional. Our new grads are brilliant, have strong and relevant training, and learn fast. Our industry veterans have built ML chips, ML compilers, and ML models, at Google or Amazon or various startups. Talent and passion makes everything work better. Several of us have previously built a chip with 30 people, where other companies might take more than 300.

Our investors

We’re backed by $25M investment from specialist investors and operators who share our vision, including:

• Daniel Gross and Nat Friedman, lead investors.
• Rajiv Khemani: Intel, Cavium, Innovium, AIspace, Auradine
• Outset Capital
• SV Angel
• Homebrew
• Amit Singh: Oracle, Google, Palo Alto Networks
• Leading LLM and AI researchers:
  • Irwan Bello
  • James Bradbury
  • Aakanksha Chowdhery
  • William Fedus
  • David Ha
• Swyx: Latent.Space, AI.Engineer, AWS, Netlify, Temporal