Skip to content

Correct stride pseudocode in p27 #240

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
dunnoconnor merged 1 commit into from
Apr 21, 2026
Merged

Correct stride pseudocode in p27 #240

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 6 additions & 2 deletions book/i18n/ko/src/puzzle_27/puzzle_27.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,10 +48,12 @@ GPU 스레드 블록 (128 스레드, 4개 또는 2개 워프, 하드웨어 조
# 복잡한 블록 전체 리덕션 (기존 방식 - Puzzle 12에서):
shared_memory[local_i] = my_value
barrier()
for stride in range(64, 0, -1):
stride = 64
while stride > 0:
if local_i < stride:
shared_memory[local_i] += shared_memory[local_i + stride]
barrier()
stride //= 2
if local_i == 0:
output[block_idx.x] = shared_memory[0]

Expand Down Expand Up @@ -81,10 +83,12 @@ if local_i == 0:
shared_memory[local_i] = my_value
barrier()
# 스트라이드 기반 인덱싱을 사용한 트리 리덕션...
for stride in range(64, 0, -1):
stride = 64
while stride > 0:
if local_i < stride:
shared_memory[local_i] += shared_memory[local_i + stride]
barrier()
stride //= 2
```

### **중간 단계: 워프 프로그래밍 (Puzzle 24)**
Expand Down
8 changes: 6 additions & 2 deletions book/src/puzzle_27/puzzle_27.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -46,10 +46,12 @@ Learn the complete parallel programming toolkit from `gpu.primitives.block`:
# Complex block-wide reduction (traditional approach - from Puzzle 12):
shared_memory[local_i] = my_value
barrier()
for stride in range(64, 0, -1):
stride = 64
while stride > 0:
if local_i < stride:
shared_memory[local_i] += shared_memory[local_i + stride]
barrier()
stride //= 2
if local_i == 0:
output[block_idx.x] = shared_memory[0]

Expand Down Expand Up @@ -79,10 +81,12 @@ Complex but educational - explicit shared memory, barriers, and tree reduction:
shared_memory[local_i] = my_value
barrier()
# Tree reduction with stride-based indexing...
for stride in range(64, 0, -1):
stride = 64
while stride > 0:
if local_i < stride:
shared_memory[local_i] += shared_memory[local_i + stride]
barrier()
stride //= 2
```

### **The intermediate step: Warp programming (Puzzle 24)**
Expand Down
Loading