Alice/my_digital_garden
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

vault backup: 2023-10-04 12:33:31

Browse source
This commit is contained in:
Alice 2023-10-04 12:33:31 +01:00
parent 47bd74f5ee
commit ce64639d15
2 changed files with 27 additions and 15 deletions
Download patch file Download diff file Expand all files Collapse all files

32
4a1s/CP/PL - Aula 4.md
View file

@ -61,21 +61,33 @@ Ganhos de 4 vezes mais.
conclusion: 20 GFlop/s
#### b)
peak with vectorization: 20 GFlop/s
peak without vectorization: 5 GFlop/s
memory bandwith limitation: 20 GFlop/s
real achievable performance:
peak with vectorization: continuous 20 GFlop/s
peak without vectorization: continuous 5 GFlop/s
memory bandwith limitation: ***see alinea d)***
real achievable performance:***see alinea c)***
measured performance:
#### c)
#### d)
memory bandwith limitation
| GFlop/s | Flop/Byte |
| ------- | --------- |
| 0.125 | |
| 0.25 | |
| 0.125 | 2.5 |
| 0.25 | 5 |
| 0.5 | 10 |
| 1 | 20 |
| 2 | |
| 4 | |
| 8 | |
| 2 | 40 |
| 4 | 80 |
| 8 | 160 |
#### c)
2 FOP (operações vírgula flutuante) -> 2 doubles (16B)
1 operation/8B -> 0.125
| GFlop/s | Flop/Byte |
| ------- | --------- |
| 0.125 | 2.5 |
#### d)
AVX -> 256b -> 4 doubles