WeiNote

yyrcd

gpu

os

pl

2020

Mar 06

2020-03-06 ·

编译器

LLVM / Clang / GCC

MLIR - Chris Lattner

MLIR (Multi-Level Intermediate Representation)

XLA - TF

XLA (Accelerated Linear Algebra) is a domain-specific compiler for linear algebra that can accelerate TensorFlow models with potentially no source code changes. Most internal benchmarks run ~1.15x faster after XLA is enabled.