NBP2P
2019-08-05 18:13:38 0 举报NBP2P 架构图
作者其他创作
大纲/内容
Non-blocking Send A_sub to p1-pn
Termination
Non-Blocking recieve A_sub
Matrix_Multiply p0's A_sub and B to generate C_sub
Compute C_serial and do comparation
Process 1-n
Wait for A_sub and B recieved
Matrix_Multiply Asub and B to generate Csub
Create random A and B
Non-blocking recieve C_sub from p1-pn and generate C
Non-Blocking recieve B
Add C_sub in p0 to C
Non-blocking Send B to p1-pn
Process 0
Blocking send C_sub to p0 (Because this is meaningless to do in non-blocking way)
Start
Wait for All requesets
0 条评论
下一页
