CIALUG December 2023 - Mojo and C/Linux intro

mojo build hello.mojo
./hello

Only amd64 and arm64 - NVIDIA GPU support Q1 2024.
JIT compilation with LLVM ORC
Multi-Level Intermediate Representation MLIR - single static assignment for all the things.
Mojo simd - n-body demo
Mojo autotune - memset demo

Performance beyond mojo

const/enum linting
GPU pin of hot cache - PowerInfer: Fast Large Language Model Serving with a Consumer-grade GPU - PowerInfer Github
tensor compression
More MILR for IO/cache levels
Equality saturation