June 2026 Community Meeting: StaMojo & MojoR

Mojo 1.0 course launch, StatMojo for statistical computing in pure Mojo, and Mojo R for 8x faster R, all from our June 2026 Modular community meeting. A course announcement plus two community talks on Mojo, GPU programming, and high-performance scientific computing. Michael opens with the first official Mojo course, launching live on YouTube in July during the Mojo 1.0 beta. The four-part series covers Mojo language fundamentals, value ownership and metaprogramming, the standard library (collections and SIMD), and a one-hour intro to GPU programming. By the end, new users can read Mojo syntax, pick up a good first issue, and start on the GPU puzzles. It is meant as a lasting learning resource for anyone getting started with Mojo. Yuhao introduces StatMojo, a young statistical computing library written in pure Mojo. It takes inspiration from SciPy's stats package and statsmodels, but stays lightweight and focused rather than covering every scientific domain. Yuhao walks through the motivation, the two-layer scope (statistical foundations plus a modeling layer with regression, GLMs, and logistic models), where the project stands today, and how to contribute. StatMojo builds on NuMojo and lives in the Mojo Math organization. Seyoon closes with Mojo R, an experiment that uses Mojo to speed up code written in R. The approach lowers R numerical kernels to Mojo, JIT compiles them, and calls the result back through R's C interface. Early benchmarks show over 8x speedups on Gibbs sampling. Seyoon covers the two-language problem, the tensor data structure that carries type, shape, and device information, and the recent move to an MLIR-style representation. He closes on a longer-term idea: Mojo and MLIR as shared infrastructure that lets R, Python, and Julia communities keep their own languages while gaining native and GPU performance. Chapters: 00:00 Welcome and agenda 00:57 Mojo 1.0 course announcement (Michael) 03:11 StatMojo: statistical computing in pure Mojo (Yuhao) 12:40 Mojo R: native and GPU performance for R (Seyoon)