Summary

This episode explores Mojo, a new language designed to unlock the full potential of modern GPUs. Ron, Chris, and Laurent discuss designing a language that balances ease of use with the necessary control for creating high-performance machine learning kernels. They delve into the trade-offs between different development environments (text editors vs. Jupyter Notebooks), the value of visualization and interactivity, and how features like borrow-checking (Rust), ref-counting (Swift & Python) can simplify machine learning workflows.

Title

Signals and Threads

Description

A technology podcast, from Jane Street. Get a peek at how Jane Street approaches technology, and how those ideas relate to the tech landscape more broadly.

Keywords

episode, street, page, systems, discuss, programming, streets, design, language, hardware, tech, machine, performance, working, learning, software, build

NS Lookup

A 3.11.250.185

Dates

Created 2026-03-09

Updated 2026-03-09

Summarized 2026-03-13