Halide_Turorial

This repository will introduce the basic of Halide -> (Algorithm + Scheduling)

Basic Functions

Tile	Tile + Fuse + Parallel	Vectorize (SIMD)	Unroll Loop
Divide whole image into tiles	Taling parallel	x86 SSE command	Reduce repeat calculation

producer(x, y) = sin(x * y)

consumer(x, y) = (producer(x, y) + producer(x, y+1) + producer(x+1, y) + producer(x+1, y+1))/4

store_root.compute_at	Tiling + compute_at
Store intermediate data in several scanlines	Divide compute_at into tiles

compute_root	compute_at
Compute all producer before use	Compute producer inside y loop

Name		Name	Last commit message	Last commit date
Latest commit History 80 Commits
figures		figures
.gitignore		.gitignore
README.md		README.md
lesson_01_basics.cpp		lesson_01_basics.cpp
lesson_02_input_image.cpp		lesson_02_input_image.cpp
lesson_03_debugging_1.cpp		lesson_03_debugging_1.cpp
lesson_04_debugging_2.cpp		lesson_04_debugging_2.cpp
lesson_05_scheduling_1.cpp		lesson_05_scheduling_1.cpp
lesson_06_realizing_over_shifted_domains.cpp		lesson_06_realizing_over_shifted_domains.cpp
lesson_07_multi_stage_pipelines.cpp		lesson_07_multi_stage_pipelines.cpp
lesson_08_Scheduling multi-stage pipelines.cpp		lesson_08_Scheduling multi-stage pipelines.cpp