issues
search
eth-cscs
/
conflux
Distributed Communication-Optimal LU-factorization Algorithm
BSD 3-Clause "New" or "Revised" License
12
stars
3
forks
source link
Optimizations
#10
Closed
kabicm
closed
3 years ago
kabicm
commented
3 years ago
This PR brings the following improvements:
Local copies:
All local copies are now multithreaded (OpenMP)
bugfix
: there were multiple bugs occuring when more than 1 thread was used.
cblas
: apart from MKL, cblas is also supported.
bugfix
: there was a memory leak with
gv.matrix
.
Optimized Bcast
: now using non-blocking MPI bcasts, where each array is requested only when it's needed.
This PR brings the following improvements:
gv.matrix
.