Validation through benchmarks over Dahu

500 x 500 over 32 cores for 100 bits over one node

Sequential

UserTime: 69.392 RealTime: 69.5579 Bitsize: 51480.7 -i 3 -q -1 -n 500 -b 100 -s 1562679163 -d "Sequential"

MPI

UserTime: 12.624 RealTime: 13.0545 Bitsize: 51477.8 -i 3 -q -1 -n 500 -b 100 -s 1562678836 -d "Distributed"

Paladin

UserTime: 278.3 RealTime: 44.9286 Bitsize: 51479.3 -i 3 -q -1 -n 500 -b 100 -s 1562678965 -d "Paladin"

500 x 500 and 100 bits among 3 nodes (only 65 cores are used) using MPI+Paladin

UserTime: 6.148 RealTime: 6.16084 Bitsize: 51479 -i 3 -q -1 -n 500 -b 100 -s 1562743984 -d "Combined"

500 x 500 and 100 bits between 2 nodes (64 cores are used) using MPI

UserTime: 13.264 RealTime: 13.5647 Bitsize: 51484.4 -i 1 -q -1 -n 500 -b 100 -s 1562742304 -d "Distributed"

1000 x 1000 and 100 bits among 3 nodes using MPI

UserTime: 97.172 RealTime: 97.4089 Bitsize: 103464 -i 3 -q -1 -n 1000 -b 100 -s 1562744869 -d "Distributed"

1000 x 1000 and 100 bits among 4 nodes(3 nodes used as workers) using MPI+Paladin

UserTime: 54.492 RealTime: 54.6705 Bitsize: 103466 -i 3 -q -1 -n 1000 -b 100 -s 1562745390 -d "Combined"