Latent Space: The AI Engineer Podcast The Shape of Compute (Chris Lattner of Modular)
949 snips
Jun 13, 2025 
Chris Lattner, the visionary behind LLVM and Swift, now leads Modular, pushing the boundaries of GPU programming with Mojo. He discusses breaking NVIDIA's hold on the market and achieving AMD performance levels. The conversation delves into Mojo's design, merging simplicity with advanced capabilities for AI applications. Lattner also shares insights on community engagement, the importance of open-source contributions, and his reflections on leadership, emphasizing the blend of technical innovation and personal balance in a startup environment.
AI Snips
Chapters
Transcript
Episode notes
Modular's Challenging R&D Journey
- Modular spent three years in a research phase focused on proving they could match NVIDIA GPU performance without CUDA.
- They achieved state-of-the-art results on LLaMA 3 models and transitioned from R&D to engineering and open source.
From CPU to GPU Transition
- Modular started with CPU optimizations and a custom compiler approach before moving to GPU support.
- Transitioning to GPU was challenging, especially without using CUDA, but it was essential to prove their technology's worth.
Persist Despite 'Impossible'
- Do not be deterred by the conventional wisdom that something is impossible; persistence is key.
- Believing in your vision and focusing on achievable milestones allows breakthrough innovation despite skepticism.