search

BoostGSoC13 / odeint-v2

odeint - parallelization GSoC project
http://headmyshoulder.github.com/odeint-v2/
0 stars 0 forks source link

MPI reduction #12

Open neapel opened 11 years ago

neapel commented 11 years ago

simple way: send data to master and reduce there; will not scale as well: O(n)

complex way: send data to node[my_index/2], reduce, repeat: O(log n).