Mentions légales du service

Skip to content
Snippets Groups Projects
Commit 3c1f2e0c authored by ANDRADE-BARROSO Guillermo's avatar ANDRADE-BARROSO Guillermo
Browse files

Update blocks_and_grids.md

parent 7ed961e1
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
CUDA/blocks_and_grids.md
+ 4
− 3
View file @ 3c1f2e0c
...@@ -95,10 +95,11 @@ Implicit copy of data from source in Host memory to GPU memory are performed wit ...@@ -95,10 +95,11 @@ Implicit copy of data from source in Host memory to GPU memory are performed wit
All the job is done by one single block of 400 threads. every thread execute a single load, multiplication and store from source to destination. This is possible because index is a bijective function of threadIdx.x that is unique index for all thread in linear block. All the job is done by one single block of 400 threads. every thread execute a single load, multiplication and store from source to destination. This is possible because index is a bijective function of threadIdx.x that is unique index for all thread in linear block.
| Data position | threadIdx.x | value of `index`| | Data position | threadIdx.x | value of `index`|
|---|---|---|---| |----|---|---|
| 0|0|0| | 0| 0| 0|
| 1|1|1| | 1| 1| 1|
| ...|...|...| | ...|...|...|
| 398|398|398| | 398|398|398|
| 399|399|399| | 399|399|399|
......
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment