[work4readme]: Split function target in two, a model dep and an independent one. Removed the sources target, which is not needed. Removed target to create dimeas traces folder.

[Issue#37]: Added target to create images from dimemas traces.

[Issue#37]: Added newcommands corresponding to parameters in param.cfg file

[Issue#37]: Changed the stream output to have a picture instead of many tables.

[Issue#37]: Added target to add dimemas images to report

[Issue#37]: Added summary of exp function to report in Makefile

[Issue#37]: Changes needed to insert multi experiments into the report

[Issue#37]: created specific target for multiExperiment targets

[Issue#37]: Update of all the receipt to create paraver legend.

[Issue#37]: Config for aggregated istruction per cycle.

[Issue#37]: Added IPC average and aggregated + MPI collective size

[Issue#37]: Added L3 total chache misses to the trace

[Issue#37]: Added cfg for L3 total chache misses

[Issue#37]: Needed to create create a scatterplot with L3 info in it

[Issue#37]: Modified in order to compute useful total L3 cm per Instruction.

[Issue#37]: plot for L3 vs IPC inserted in json target

[Issue#37]: STREAM test moved to target

[Issue#37]: Moved *.dim traces to dimemas_trace folder. Fixed name of paramedir and dimemas target

[Issue#37]: Started using native Makefile variables more aggressively

[Issue#37]: Trying to move all the bash in paramedir target to file

[Issue#37]: Find a new useful Makefile automatic var, plus moved paramedir to target

[Issue#37]: paraver target

[Issue#37]: Added absolute path to cfg, fixed some bugs

[Issue#37]: added option to set slurm time in seconds

[Issue#37]: prv2dim process has to be moved to slurm

[Issue#37]: Added target for preproc/clean_preproc report

[Issue#37]: Paraver moved to a unique target, now creates a folder for all the images, and creates legend.

[Issue#37]: Added cfg for computing nemo efficiency. Weakscaling tests is not possible, since Orca025 data are not on Zenodo

[Issue#37]: The name of the branch is sometimes a mess, and latex does not like it, for the moment the name of the model will do

sparonuz authored Aug 04, 2020 and sparonuz committed Aug 05, 2020

Files can be used for captions

Added index. This requires the .tex to be compiled twice, since the first produce the table of contents, and the second adds it to the doc

otintopr authored Jul 10, 2020 and sparonuz committed Jul 31, 2020

added script to load some of the needed modules, and corresponding check in makefile

Added functio to purge csv from empty columns. Moved OptParse --> ArgParse

Text to be added as description of the benchmarks run

report command taken out of targets

otintopr authored Jul 10, 2020 and sparonuz committed Jul 30, 2020

Added script to load some of the needed modules, and corresponding check in makefile

Added functio to purge csv from empty columns. Moved OptParse --> ArgParse

Added hard placing of fig and tabs

Some files were saved on disk with same name, thus overwriting others.

Handled case of empty lines in csv file

target to process multi experiments has been added

Changed multiexperiment_comparison to compute SYPD showing nodes,and not cores, so as to be used for affinity study

Moved all config for multi experiment to one folder, added cfg for affinity

Added files to test affinity of BENCH test and NEMO

Split experiment recipe btw json and scalability, which is the thing needed for affinity.

Example on how to analize affinity results and add it to report

Issue#16

See merge request ces/nemo_performance_profile!7

Issue#16

See merge request ces/nemo_performance_profile!6