chn/bscpkgs
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
737 Commits 1 Branch 0 Tags 1.8 MiB
83921d1788
Commit Graph

192 Commits

Author SHA1 Message Date
Rodrigo Arias Mallo
f8122f3c8b heat: use the hcut tool to limit the cpus 2021-04-06 11:05:56 +02:00
Rodrigo Arias Mallo
d68ce914ba heat: use cut to partition the trace 
The awk script doesn't take in consideration the events close to the cut
points, which are significative with low parallelism.
 2021-04-06 11:05:51 +02:00
Rodrigo Arias Mallo
cb482fa3ea heat: remove perf from the ctf experiment 
As we would be extracting perf stats from the trace processing steps.
 2021-04-06 11:05:10 +02:00
Rodrigo Arias Mallo
10b1ff8f7a saiph: simplify granularity and ss experiments 2021-04-01 19:25:38 +02:00
Sandra
b64b864194 saiph: clean exps and figs 2021-04-01 19:25:37 +02:00
Sandra
46536548ca saiph: update scaling exp and figures 2021-04-01 19:24:38 +02:00
Sandra
5caf2f79f3 saiph: change scaling R script 2021-04-01 19:24:38 +02:00
Sandra
a90c044c3e saiph: add manual global blocking 
Ensure cuts in a single dimension
 2021-04-01 19:24:38 +02:00
Sandra
ddef901e2f saiph: add nsteps parameter to experiments 2021-03-30 18:54:35 +02:00
Rodrigo Arias Mallo
d108306a29 saiph: add blocking experiments to index 
Remove unused environment variables as well.
 2021-03-30 18:54:35 +02:00
Sandra
37e11c749f saiph: add cacheline compilation parameter 2021-03-30 18:54:35 +02:00
Sandra
02a62c18ac saiph: add strong scaling experiment 2021-03-30 18:54:35 +02:00
Sandra
a2306eb941 saiph: add some blocking experiments 2021-03-30 18:54:35 +02:00
Sandra
38d4d0b48c saiph: delete extrae XML configuration files 2021-03-30 18:54:35 +02:00
Rodrigo Arias Mallo
992af14c7f saiph: add scaling experiment 2021-03-30 18:54:35 +02:00
Rodrigo Arias Mallo
a4b2dfddb4 saiph: update granularity experiment 2021-03-30 18:54:35 +02:00
Rodrigo Arias Mallo
830d648925 saiph: reduce the number of loops 
The current app Heat3D_vect has a long initialization time
 2021-03-30 18:54:16 +02:00
Rodrigo Arias Mallo
e4ab177d6c saiph: remove dangerous Intel MPI envvar 
It is no longer used, as we have moved to the release library version.
 2021-03-30 17:56:26 +02:00
Rodrigo Arias Mallo
5ac581b573 creams: remove pure mpi from granularity 2021-03-30 16:14:32 +02:00
Rodrigo Arias Mallo
b900cb95f0 creams: make configurations unique 2021-03-30 16:14:11 +02:00
Rodrigo Arias Mallo
389d3f6310 creams: simplify granularity figure 2021-03-30 16:07:14 +02:00
Rodrigo Arias Mallo
87f751185c creams: merge similar experiments together 
Large experiments have the enableExtended parameter disabled by default,
which enables more tests.
 2021-03-30 15:55:57 +02:00
Pedro Martinez
5cd9894636 creams: redefine granularity experiments 2021-03-24 13:52:26 +01:00
Pedro Martinez
bfc32ef4b7 creams: readjust granularity for strong scalability 2021-03-24 13:52:26 +01:00
Pedro Martinez
cb4d27aefb creams: bugfix in granularity values 2021-03-24 13:52:26 +01:00
Pedro Martinez
d27c696259 creams: reduce granularity combinations to 8 2021-03-24 13:52:26 +01:00
Pedro Martinez
a55019c6ef creams: add more nodes for granularity experiments 2021-03-24 13:52:26 +01:00
Rodrigo Arias Mallo
c59f298ae2 creams: reduce granularity experiment units 2021-03-24 13:52:26 +01:00
Rodrigo Arias Mallo
6818b29d02 creams: fix outdated nanos6.toml 
This temporal fix allows the experiment to ignore the nanos6.toml in the
git repository, and only set version.dependencies variable.
 2021-03-24 13:52:26 +01:00
Pedro Martinez
8445fb0928 creams: run the cp command in one process only 2021-03-24 13:52:26 +01:00
Pedro Martinez
1aa0e77157 creams: avoid race condition 
Ensure only one Slurm process performs environment operations
 2021-03-24 13:52:26 +01:00
Pedro Martinez
938246322f creams: add OpenMP branches 2021-03-24 13:52:26 +01:00
Pedro Martinez
6c0f4ec1b3 creams: add granularity experiments 2021-03-24 13:52:26 +01:00
Rodrigo Arias Mallo
c41456412c examples: Add granularity examples 2021-03-12 19:33:40 +01:00
Rodrigo Arias Mallo
7d66b34140 nbody: fix converter rename in nanos6 CTF options 2021-03-12 12:58:41 +01:00
Rodrigo Arias Mallo
0781e8b28e nbody: remove jemalloc experiments 
Nanos6 has jemalloc enabled by default
 2021-03-12 12:58:41 +01:00
Rodrigo Arias Mallo
637c57b388 nbody: improve unit name 2021-03-12 12:58:41 +01:00
Rodrigo Arias Mallo
3a2694ad36 nbody: add mpi branch in scaling experiment 2021-03-12 12:58:37 +01:00
Rodrigo Arias Mallo
425479c9fc nbody: add scaling experiment 2021-03-12 12:57:01 +01:00
Rodrigo Arias Mallo
854707103c heat: add ctf stage to analyze mode times 2021-03-12 12:13:10 +01:00
Rodrigo Arias Mallo
b192fc44f5 heat: refactor cache into granul experiment 2021-03-09 18:45:33 +01:00
Rodrigo Arias Mallo
b600f64fcc heat: add cache miss experiment and figure 2021-03-05 18:31:31 +01:00
Rodrigo Arias Mallo
c1efba1e65 heat: rename test -> granul experiment 2021-03-05 18:28:32 +01:00
Rodrigo Arias Mallo
363700eb9a heat: update test experiment 2021-03-05 16:18:51 +01:00
Rodrigo Arias Mallo
6973f48638 osu: add an experiment for Intel MPI tunning 2021-03-03 12:42:19 +01:00
Rodrigo Arias Mallo
ed932c9921 osu: add bw test 2021-03-01 11:58:23 +01:00
Rodrigo Arias Mallo
a36d912022 osu: add multithread benchmark 2021-03-01 11:55:13 +01:00
Rodrigo Arias Mallo
ceb25e5d18 osu: add figure for latency tests 2021-02-23 17:52:48 +01:00
Rodrigo Arias Mallo
0c9e89dcc0 osu: update experiments using stdexp 2021-02-23 15:22:56 +01:00
Rodrigo Arias Mallo
9beda65778 lulesh: add experiment with all variants 2021-01-28 15:27:21 +01:00
Rodrigo Arias Mallo
892fb35d27 nbody: Fix infinite recursion 
We want to override the previous layer (super), not the last one (self).
 2021-01-11 14:30:12 +01:00
Pedro Martinez
203dc9f295 Configure the nanos6 environment and get the right hardware attributes 2020-12-15 19:34:49 +01:00
Pedro Martinez
2e18761b48 Use the 'hw' attributes 2020-12-15 19:11:29 +01:00
Pedro Martinez
748d335a39 Define variables 'ntasksPerNode' and 'cpusPerTask' for each experiment and other minor changes 2020-12-15 19:11:29 +01:00
Rodrigo Arias Mallo
9a0ea08d72 Reorganization 
- All garlic stuff is moved into garlic/
- Group the overlay index by sections
- Add a garlic/default.nix link to the main default.nix, so we can
  build derivations at garlic/
 2020-12-07 13:33:42 +01:00
Rodrigo Arias Mallo
bdaadd4ef7 nbody: add ctf tests 2020-12-03 13:20:40 +01:00
Rodrigo Arias Mallo
aca7e36fc7 bigsort: add experiment with input generation 2020-11-20 15:41:27 +01:00
Rodrigo Arias Mallo
4f0da10321 bigsort: Use cpusPerTask instead of cpuBind 2020-11-20 13:57:12 +01:00
David Alvarez
0c438d4dac Setup for test experiment 2020-11-20 13:57:12 +01:00
David Alvarez
a0dac209e3 First test experiment 2020-11-20 13:57:12 +01:00
Rodrigo Arias Mallo
fe0bd8b200 creams: fix pure experiment 
Use machine agnostic specification for resources
 2020-11-17 12:31:03 +01:00
Rodrigo Arias Mallo
016422cede Update nbody experiment 
Generate the input based on the target machine description.
 2020-11-17 11:26:35 +01:00
Rodrigo Arias Mallo
5e50ef19fe Update experiments with cpusPerTask 
Try to avoid manually setting the hardware specs and rather use
the hw attrset.
 2020-11-17 11:17:57 +01:00
Rodrigo Arias Mallo
74537e682c Use divisors in the slurm cpu experiment 2020-11-17 11:01:34 +01:00
Rodrigo Arias Mallo
dea523460a Add slurm affinity experiment 2020-11-17 11:01:34 +01:00
Sandra
4ae66adb9a Saiph: adding granularity experiment and figures 2020-11-13 09:56:40 +01:00
Sandra
86d1d426ec Saiph: Removing devMode parameter 2020-11-12 19:10:43 +01:00
Rodrigo Arias Mallo
966606b62d hpcg: precompute the input 2020-11-09 17:48:46 +01:00
Rodrigo Arias Mallo
11ac02da08 heat: Add test experiment and plot 2020-11-05 19:56:26 +01:00
Rodrigo Arias Mallo
074a75facb saiph: name the experiment and units in numcomm 2020-11-05 19:53:38 +01:00
Rodrigo Arias Mallo
9e477a2313 hpcg: smaller input size 2020-11-05 19:46:34 +01:00
Rodrigo Arias Mallo
0c58bb63b5 hpcg: add exp and unit name 2020-11-03 19:10:00 +01:00
Rodrigo Arias Mallo
5eea48c5b0 Add exp and unit name to nbody tampi experiment 2020-11-03 19:10:00 +01:00
Rodrigo Arias Mallo
d5d42b3c09 Add unit and exp name to nbody test 2020-11-03 19:10:00 +01:00
Rodrigo Arias Mallo
a66cdb52fb nbody: Fix test experiment 2020-11-03 19:10:00 +01:00
Rodrigo Arias Mallo
3bd4e61f3f WIP: Testing with automatic fetching 2020-11-03 19:09:59 +01:00
Rodrigo Arias Mallo
8ce88ef046 Add dataset attrset in garlic 
Modify nbody to evenly distribute blocks per cpu
 2020-11-03 19:09:59 +01:00
Rodrigo Arias Mallo
06c29b573f Add exp.nbody.tampi variants 2020-11-03 19:09:59 +01:00
Rodrigo Arias Mallo
4beb069627 WIP: postprocessing pipeline 
Now each run is executed in a independent folder
 2020-11-03 19:09:59 +01:00
Rodrigo Arias Mallo
1321b6a888 Add experiments with jemalloc and CPU affinity 2020-11-03 19:09:59 +01:00
Rodrigo Arias Mallo
308673f7f6 Increase nbody test cases 2020-11-03 19:09:59 +01:00
Rodrigo Arias Mallo
65745e0aaf WIP: Add another nbody experiment 2020-11-03 19:09:58 +01:00
Rodrigo Arias Mallo
72ba080db1 Merge branch 'nbody' into 'master' 
NBody Experiments

See merge request rarias/bscpkgs!5
 2020-10-30 16:25:27 +01:00
Raúl Peñacoba
9c20537f91 Since mpi+omp version uses 6 threads, change nblocks values 2020-10-30 15:11:23 +01:00
Raúl Peñacoba
6a1375726f Fix problem sizes to be equivalent between versions 2020-10-30 14:44:33 +01:00
Raúl Peñacoba
d757332448 Remove extrae home 2020-10-30 14:35:09 +01:00
Raúl Peñacoba
58e3d48a16 Use mask_cpu and n.x n.y n.z instead of n 2020-10-30 14:08:56 +01:00
Raúl Peñacoba
b856e2147a Use discrete deps in nanos6. Pass nblocks to omp version and use the same experiments as oss 2020-10-30 14:08:56 +01:00
Raúl Peñacoba
22a294f9cc Forgot to set one task per node 2020-10-30 14:08:55 +01:00
Raúl Peñacoba
ea0272c212 Add OmpSs-2 (no mpi) version 2020-10-30 14:08:55 +01:00
Raúl Peñacoba
e20061254b WIP: Add mpi, omp and mpi+omp experiments. See more. 
Seems that gcc compilation with OpenMP throws an error. Investigate.
I think I've forgot to add an override of mpicxx compiler backend
 2020-10-30 14:08:55 +01:00
Rodrigo Arias Mallo
01b2584688 Update hpcg experiments 2020-10-30 14:08:55 +01:00
Raúl Peñacoba
7bf3e81233 WIP: trying to make mpi branch working 2020-10-30 14:08:55 +01:00
Raúl Peñacoba
6bd7e12cff WIP: forgot to add the folder 2020-10-30 14:08:55 +01:00
Antoni Navarro
05ce36e158 Add the MPI-weak scaling experiment and strong scaling experiments 2020-10-29 16:31:21 +01:00
Antoni Navarro
6ccc159487 Fix one of the CPU Masks in the weak scaling experiment 2020-10-29 16:30:55 +01:00
Antoni Navarro
8b985de65d Add a few scalability experiments for some variants 2020-10-28 15:35:09 +01:00
Rodrigo Arias Mallo
aa1ffa5208 Remove unused experiments 2020-10-09 20:17:35 +02:00
Rodrigo Arias Mallo
1b703bd431 Fix saiph numcomm experiment 2020-10-09 19:40:49 +02:00
Rodrigo Arias Mallo
298c7362b3 New config design 2020-10-09 19:33:06 +02:00