0. Intro to Supermini: A Mac Mini Supercomputer
In this series
What is this?
A supercomputer made out of Intel Mac Minis. It's running the typical HPC Linux stack:
- Linux
- NFS
- Slurm
- Lmod
- MPI
Images
Linpack results
5 nodes hit 1.313 Tflops, scaling factor of 4.366 over single node.
This run was done on: Tue Oct 8 04:37:12 PM EDT 2024
RANK=0, NODE=0-0
RANK=4, NODE=0-0
RANK=1, NODE=0-0
RANK=3, NODE=0-0
RANK=2, NODE=0-0
Number of Heterogeneous data entries for HPL.dat=0
================================================================================
HPLinpack 2.3 -- High-Performance Linpack benchmark -- December 2, 2018
Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
Modified by Julien Langou, University of Colorado Denver
================================================================================
An explanation of the input/output parameters follows:
T/V : Wall time / encoded variant.
N : The order of the coefficient matrix A.
NB : The partitioning blocking factor.
P : The number of process rows.
Q : The number of process columns.
Time : Time in seconds to solve the linear system.
Gflops : Rate of execution for solving the linear system.
The following parameter values will be used:
N : 130944
NB : 192
PMAP : Row-major process mapping
P : 5
Q : 1
PFACT : Right
NBMIN : 4
NDIV : 2
RFACT : Crout
BCAST : 1ringM
DEPTH : 1
SWAP : Mix (threshold = 64)
L1 : transposed form
U : transposed form
EQUIL : yes
ALIGN : 8 double precision words
--------------------------------------------------------------------------------
- The matrix A is randomly generated for each test.
- The following scaled residual check will be computed:
||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
- The relative machine precision (eps) is taken to be 1.110223e-16
- Computational tests pass if scaled residuals are less than 16.0
^C^CDone: Tue Oct 8 04:37:34 PM EDT 2024
[noah@supermini mp_linpack]$ ^C
[noah@supermini mp_linpack]$ ^C
[noah@supermini mp_linpack]$ fg
bash: fg: current: no such job
(reverse-i-search)`emacs': ^Cacs HPL.dat
[noah@supermini mp_linpack]$ emacs runme_intel64_dynamic
[noah@supermini mp_linpack]$ ./runme_intel64_dynamic
This is a SAMPLE run script. Change it to reflect the correct number
of CPUs/threads, number of nodes, MPI processes per node, etc..
This run was done on: Tue Oct 8 04:38:16 PM EDT 2024
RANK=3, NODE=0-0
RANK=4, NODE=0-0
RANK=1, NODE=0-0
RANK=2, NODE=0-0
RANK=0, NODE=0-0
Number of Heterogeneous data entries for HPL.dat=0
================================================================================
HPLinpack 2.3 -- High-Performance Linpack benchmark -- December 2, 2018
Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
Modified by Julien Langou, University of Colorado Denver
================================================================================
An explanation of the input/output parameters follows:
T/V : Wall time / encoded variant.
N : The order of the coefficient matrix A.
NB : The partitioning blocking factor.
P : The number of process rows.
Q : The number of process columns.
Time : Time in seconds to solve the linear system.
Gflops : Rate of execution for solving the linear system.
The following parameter values will be used:
N : 130944
NB : 192
PMAP : Row-major process mapping
P : 5
Q : 1
PFACT : Right
NBMIN : 4
NDIV : 2
RFACT : Crout
BCAST : 1ringM
DEPTH : 1
SWAP : Mix (threshold = 64)
L1 : transposed form
U : transposed form
EQUIL : yes
ALIGN : 8 double precision words
--------------------------------------------------------------------------------
- The matrix A is randomly generated for each test.
- The following scaled residual check will be computed:
||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
- The relative machine precision (eps) is taken to be 1.110223e-16
- Computational tests pass if scaled residuals are less than 16.0
mini3 : Column=000768 Fraction=0.005 Kernel= 0.85 Mflops=1787779.96
mini1 : Column=001344 Fraction=0.010 Kernel=1429562.53 Mflops=1615337.81
mini5 : Column=002112 Fraction=0.015 Kernel=1270621.32 Mflops=1471649.09
mini3 : Column=002688 Fraction=0.020 Kernel=1365993.42 Mflops=1448042.15
mini2 : Column=003456 Fraction=0.025 Kernel=1370329.92 Mflops=1430396.00
mini5 : Column=004032 Fraction=0.030 Kernel=1367323.40 Mflops=1421272.13
mini3 : Column=004608 Fraction=0.035 Kernel=1324791.88 Mflops=1408850.37
mini2 : Column=005376 Fraction=0.040 Kernel=1314294.12 Mflops=1395029.61
mini5 : Column=005952 Fraction=0.045 Kernel=1314519.96 Mflops=1387143.21
mini4 : Column=006720 Fraction=0.050 Kernel=1366459.27 Mflops=1384869.91
mini2 : Column=007296 Fraction=0.055 Kernel=1394087.78 Mflops=1385551.91
mini5 : Column=007872 Fraction=0.060 Kernel=1356850.97 Mflops=1383524.74
mini4 : Column=008640 Fraction=0.065 Kernel=1377632.06 Mflops=1383042.34
mini2 : Column=009216 Fraction=0.070 Kernel=1362021.06 Mflops=1381796.10
mini1 : Column=009984 Fraction=0.075 Kernel=1394195.75 Mflops=1382675.39
mini4 : Column=010560 Fraction=0.080 Kernel=1365994.88 Mflops=1381829.13
mini2 : Column=011136 Fraction=0.085 Kernel=1381746.20 Mflops=1381821.41
mini1 : Column=011904 Fraction=0.090 Kernel=1373720.01 Mflops=1381343.77
mini4 : Column=012480 Fraction=0.095 Kernel=1371374.71 Mflops=1380926.03
mini3 : Column=013248 Fraction=0.100 Kernel=1369656.59 Mflops=1380323.49
mini1 : Column=013824 Fraction=0.105 Kernel=1375464.00 Mflops=1380148.81
mini5 : Column=014592 Fraction=0.110 Kernel=1380493.67 Mflops=1380156.23
mini3 : Column=015168 Fraction=0.115 Kernel=1355009.47 Mflops=1379298.50
mini1 : Column=015744 Fraction=0.120 Kernel=1376992.68 Mflops=1379231.33
mini5 : Column=016512 Fraction=0.125 Kernel=1362429.38 Mflops=1378531.31
mini3 : Column=017088 Fraction=0.130 Kernel=1364105.33 Mflops=1378106.24
mini2 : Column=017856 Fraction=0.135 Kernel=1359921.12 Mflops=1377427.46
mini5 : Column=018432 Fraction=0.140 Kernel=1359567.99 Mflops=1376936.60
mini3 : Column=019008 Fraction=0.145 Kernel=1378247.04 Mflops=1376971.54
mini2 : Column=019776 Fraction=0.150 Kernel=1364184.35 Mflops=1376552.37
mini5 : Column=020352 Fraction=0.155 Kernel=1367872.74 Mflops=1376338.40
mini4 : Column=021120 Fraction=0.160 Kernel=1361219.89 Mflops=1375884.11
mini2 : Column=021696 Fraction=0.165 Kernel=1354239.97 Mflops=1375395.72
mini5 : Column=022272 Fraction=0.170 Kernel=1360163.68 Mflops=1375061.35
mini4 : Column=023040 Fraction=0.175 Kernel=1372901.97 Mflops=1375012.40
mini2 : Column=023616 Fraction=0.180 Kernel=1348560.72 Mflops=1374473.31
mini1 : Column=024384 Fraction=0.185 Kernel=1358558.72 Mflops=1374065.33
mini4 : Column=024960 Fraction=0.190 Kernel=1359095.70 Mflops=1373787.95
mini2 : Column=025536 Fraction=0.195 Kernel=1344261.10 Mflops=1373241.57
mini1 : Column=026304 Fraction=0.200 Kernel=1366601.59 Mflops=1373089.33
mini4 : Column=026880 Fraction=0.205 Kernel=1345970.35 Mflops=1372626.72
mini3 : Column=027648 Fraction=0.210 Kernel=1347525.50 Mflops=1372063.40
mini1 : Column=028224 Fraction=0.215 Kernel=1369505.07 Mflops=1372029.78
mini5 : Column=028992 Fraction=0.220 Kernel=1344479.87 Mflops=1371450.56
mini3 : Column=029568 Fraction=0.225 Kernel=1350248.31 Mflops=1371131.33
mini1 : Column=030144 Fraction=0.230 Kernel=1355506.80 Mflops=1370909.63
mini5 : Column=030912 Fraction=0.235 Kernel=1356098.62 Mflops=1370623.06
mini3 : Column=031488 Fraction=0.240 Kernel=1333029.18 Mflops=1370096.85
mini2 : Column=032256 Fraction=0.245 Kernel=1352388.21 Mflops=1369785.93
mini5 : Column=032832 Fraction=0.250 Kernel=1350826.08 Mflops=1369533.52
mini3 : Column=033408 Fraction=0.255 Kernel=1330639.27 Mflops=1369031.31
mini2 : Column=034176 Fraction=0.260 Kernel=1316447.75 Mflops=1368148.77
mini5 : Column=034752 Fraction=0.265 Kernel=1249522.82 Mflops=1366603.87
mini4 : Column=035520 Fraction=0.270 Kernel=1249341.62 Mflops=1364645.16
mini2 : Column=036096 Fraction=0.275 Kernel=1226975.53 Mflops=1362922.38
mini5 : Column=036672 Fraction=0.280 Kernel=1198126.08 Mflops=1360862.61
mini4 : Column=037440 Fraction=0.285 Kernel=1230691.81 Mflops=1358827.45
mini2 : Column=038016 Fraction=0.290 Kernel=1209136.59 Mflops=1357074.32
mini1 : Column=038784 Fraction=0.295 Kernel=1223142.35 Mflops=1355073.86
mini4 : Column=039360 Fraction=0.300 Kernel=1216037.96 Mflops=1353548.71
mini3 : Column=040128 Fraction=0.305 Kernel=1234050.98 Mflops=1351862.59
mini1 : Column=040704 Fraction=0.310 Kernel=1230583.99 Mflops=1350621.09
mini4 : Column=041280 Fraction=0.315 Kernel=1152487.86 Mflops=1348497.77
mini3 : Column=042048 Fraction=0.320 Kernel=1221820.36 Mflops=1346825.94
mini1 : Column=042624 Fraction=0.325 Kernel=1236800.73 Mflops=1345788.77
mini5 : Column=043392 Fraction=0.330 Kernel=1285457.08 Mflops=1345068.17
mini3 : Column=043968 Fraction=0.335 Kernel=1306660.43 Mflops=1344740.09
mini1 : Column=044544 Fraction=0.340 Kernel=1209770.83 Mflops=1343530.25
mini5 : Column=045312 Fraction=0.345 Kernel=1292373.39 Mflops=1342964.06
mini3 : Column=045888 Fraction=0.350 Kernel=1307699.85 Mflops=1342684.00
mini2 : Column=046656 Fraction=0.355 Kernel=1315406.47 Mflops=1342409.46
mini5 : Column=047232 Fraction=0.360 Kernel=1307794.20 Mflops=1342144.05
mini3 : Column=047808 Fraction=0.365 Kernel=1258302.50 Mflops=1341500.08
mini2 : Column=048576 Fraction=0.370 Kernel=1238980.60 Mflops=1340467.09
mini5 : Column=049152 Fraction=0.375 Kernel=1354814.07 Mflops=1340560.33
mini4 : Column=049920 Fraction=0.380 Kernel=1305963.70 Mflops=1340252.79
mini2 : Column=050496 Fraction=0.385 Kernel=1332320.94 Mflops=1340195.97
mini5 : Column=051072 Fraction=0.390 Kernel=1327935.81 Mflops=1340112.70
mini4 : Column=051840 Fraction=0.395 Kernel=1290520.60 Mflops=1339693.34
mini2 : Column=052416 Fraction=0.400 Kernel=1213642.16 Mflops=1338840.90
mini1 : Column=053184 Fraction=0.405 Kernel=1283695.21 Mflops=1338386.79
mini4 : Column=053760 Fraction=0.410 Kernel=1284763.42 Mflops=1338065.55
mini3 : Column=054528 Fraction=0.415 Kernel=1228494.63 Mflops=1337155.16
mini1 : Column=055104 Fraction=0.420 Kernel=1231417.09 Mflops=1336526.21
mini4 : Column=055680 Fraction=0.425 Kernel=1297392.64 Mflops=1336310.91
mini3 : Column=056448 Fraction=0.430 Kernel=1267898.00 Mflops=1335793.10
mini1 : Column=057024 Fraction=0.435 Kernel=1279975.08 Mflops=1335498.45
mini5 : Column=057792 Fraction=0.440 Kernel=1225148.64 Mflops=1334684.92
mini3 : Column=058368 Fraction=0.445 Kernel=1315977.21 Mflops=1334590.72
mini1 : Column=058944 Fraction=0.450 Kernel=1295652.52 Mflops=1334403.67
mini5 : Column=059712 Fraction=0.455 Kernel=1308077.56 Mflops=1334229.79
mini3 : Column=060288 Fraction=0.460 Kernel=1337913.58 Mflops=1334245.82
mini2 : Column=061056 Fraction=0.465 Kernel=1319032.40 Mflops=1334158.52
mini5 : Column=061632 Fraction=0.470 Kernel=1314428.96 Mflops=1334067.08
mini3 : Column=062208 Fraction=0.475 Kernel=1296426.15 Mflops=1333900.22
mini2 : Column=062976 Fraction=0.480 Kernel=1302861.42 Mflops=1333727.75
mini5 : Column=063552 Fraction=0.485 Kernel=1341225.19 Mflops=1333754.55
mini4 : Column=064320 Fraction=0.490 Kernel=1307071.12 Mflops=1333618.67
mini2 : Column=064896 Fraction=0.495 Kernel=1286874.29 Mflops=1333425.58
mini1 : Column=067584 Fraction=0.515 Kernel=1307230.85 Mflops=1332974.18
mini4 : Column=070080 Fraction=0.535 Kernel=1294758.57 Mflops=1332414.89
mini3 : Column=072768 Fraction=0.555 Kernel=1291716.09 Mflops=1331820.04
mini2 : Column=075456 Fraction=0.575 Kernel=1285163.87 Mflops=1331217.73
mini5 : Column=077952 Fraction=0.595 Kernel=1279205.02 Mflops=1330645.61
mini4 : Column=080640 Fraction=0.615 Kernel=1274748.12 Mflops=1330061.48
mini3 : Column=083328 Fraction=0.635 Kernel=1261108.99 Mflops=1329394.36
mini1 : Column=085824 Fraction=0.655 Kernel=1255072.03 Mflops=1328812.15
mini5 : Column=088512 Fraction=0.675 Kernel=1238588.02 Mflops=1328117.57
mini3 : Column=091008 Fraction=0.695 Kernel=1225553.97 Mflops=1327469.40
mini2 : Column=104256 Fraction=0.795 Kernel=1187356.49 Mflops=1324337.03
mini5 : Column=117312 Fraction=0.895 Kernel=836359.61 Mflops=1318681.71
mini3 : Column=130368 Fraction=0.995 Kernel=301358.90 Mflops=1313677.12
================================================================================
T/V N NB P Q Time Gflops
--------------------------------------------------------------------------------
WR11C2R4 130944 192 5 1 1139.96 1.31306e+03
HPL_pdgesv() start time Tue Oct 8 16:38:20 2024
HPL_pdgesv() end time Tue Oct 8 16:57:20 2024
--------------------------------------------------------------------------------
||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 4.36323756e-03 ...... PASSED
================================================================================
Finished 1 tests with the following results:
1 tests completed and passed residual checks,
0 tests completed and failed residual checks,
0 tests skipped because of illegal input values.
--------------------------------------------------------------------------------
End of Tests.
================================================================================
+ echo -n 'Done: '
+ date
+ echo -n 'Done: '
Done: + date
Tue Oct 8 04:57:21 PM EDT 2024