AI Infrastructure SoftwareChris Lattner et al. (Google)Google

Google presents MLIR

Presented MLIR — reusable, extensible compiler infrastructure spanning multiple levels of abstraction

Why it matters

The common foundation shared by XLA, TVM, Triton and Mojo. Ended the practice of writing a new compiler for every accelerator

Source — Primary sources first; secondary reports are labelled “Press”

Paperarxiv.orghttps://arxiv.org/abs/2002.11054

Added in Week 35, 2026 · See that week

Nearby on the same axis

Microsoft Research open-sources DeepSpeedOfficialPaszke et al. publish the PyTorch paperPaperMicrosoft presents ZeROPaperNVIDIA presents Megatron-LMPaper