marenostrum.md

Compilation problems on MareNostrum cluster

I'll keep the list updated as I go.

1. The installed version of boost is 1.64, which has a bug in boost.serialization, resolved by this patch. An alternative solution is to do

   #include <boost/serialization/array_wrapper.hpp>

   before including any boost headers.
2. Link-time errors caused by the fact that the installed boost was compiled with pre-C++11 std::basic_string, whereas the g++ v7 compiler has C++11 standard enabled by default. This may be cured by -D_GLIBCXX_USE_CXX11_ABI=0 preprocessor definition (see this SO question).

omp-scalability.md

$ lscpu
lscpu 
Architecture:          x86_64
CPU op-mode(s):        32-bit, 64-bit
Byte Order:            Little Endian
CPU(s):                48
On-line CPU(s) list:   0-47
Thread(s) per core:    1
Core(s) per socket:    24
Socket(s):             2
NUMA node(s):          2
Vendor ID:             GenuineIntel
CPU family:            6
Model:                 85
Model name:            Intel(R) Xeon(R) Platinum 8160 CPU @ 2.10GHz
Stepping:              4
CPU MHz:               2100.000
CPU max MHz:           2101.0000
CPU min MHz:           1000.0000
BogoMIPS:              4190.15
Virtualization:        VT-x
L1d cache:             32K
L1i cache:             32K
L2 cache:              1024K
L3 cache:              33792K
NUMA node0 CPU(s):     0-23
NUMA node1 CPU(s):     24-47
Flags:                 fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch ida arat epb pln pts dtherm intel_pt tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm cqm mpx avx512f avx512dq rdseed adx smap clflushopt clwb avx512cd avx512bw avx512vl xsaveopt xsavec xgetbv1 cqm_llc cqm_occup_llc

2D Poisson

$ for n in 1 2 4 6 8 16 24 48; do
$   echo --- ${n} ---
$   OMP_PLACES=sockets OMP_NUM_THREADS=${n} ./solver -n 200 | grep -A1 setup
$ done
--- 1 ---
[  setup:         13.624 s] ( 36.76%)
[  solve:         22.397 s] ( 60.42%)
--- 2 ---
[  setup:          7.832 s] ( 37.09%)
[  solve:         12.643 s] ( 59.88%)
--- 4 ---
[  setup:          4.874 s] ( 40.92%)
[  solve:          6.417 s] ( 53.87%)
--- 6 ---
[  setup:          3.401 s] ( 39.23%)
[  solve:          4.310 s] ( 49.72%)
--- 8 ---
[  setup:          3.017 s] ( 41.75%)
[  solve:          3.254 s] ( 45.03%)
--- 16 ---
[  setup:          2.217 s] ( 44.74%)
[  solve:          1.786 s] ( 36.04%)
--- 24 ---
[  setup:          2.206 s] ( 51.54%)
[  solve:          1.461 s] ( 34.15%)
--- 48 ---
[  setup:          2.142 s] ( 43.20%)
[  solve:          1.849 s] ( 37.28%)

3D Tube (Navier-Stokes problem)

$ for n in 1 2 4 8 16 24 48; do
$   echo --- ${n} ---
$   OMP_PLACES=sockets OMP_NUM_THREADS=${n} ./block4_schur -A tube3d/A.bin -f tube3d/b.bin | grep -A1 setup
$ done