Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add logs and summary tables of a DY+3j subprocess for CHEP24 #1024

Closed
wants to merge 46 commits into from

Conversation

valassi
Copy link
Member

@valassi valassi commented Oct 12, 2024

Hi @oliviermattelaer this PR is only adding logs and summary tables of a DY+3j subprocess for CHEP24.

Hopefully not controversial. The only change in CODEGEN is the addition of a reference logfile. This includes the earlier #1021.

Can you please have a look? Thanks Andrea

PS Example
https://github.com/madgraph5/madgraph4gpu/blob/7b9201b67dda6cf680cc0af80f83ed796047a3ef/epochX/cudacpp/tmad/summaryTable_dy3j_chep24.txt

*** FPTYPE=m ******************************************************************

+++  REVISION ca715b5 (commit date: 2024-10-12 12:55:17 +0200) +++
On itgold91.cern.ch [CPU: Intel(R) Xeon(R) Gold 6326 CPU] [GPU: none]:

===========================================================================================================
|            | mad                        | mad               | mad               | sa/brdg   | sa/full   |
-----------------------------------------------------------------------------------------------------------
| guxtaptamggux  | [sec] tot = mad + MEs  | [TOT/sec]         | [MEs/sec]         | [MEs/sec] | [MEs/sec] |
===========================================================================================================
| nevt/grid  |                            |                   |                   |      8192 |      8192 |
| nevt total |                      81920 |             81920 |             81920 |  256*32*1 |  256*32*1 |
-----------------------------------------------------------------------------------------------------------
| FORTRAN    |   37.42 =  11.44 +   25.98 |  2.19e+03 (= 1.0) |  3.15e+03 (= 1.0) |       --- |       --- |
| CPP/none   |   37.14 =  11.46 +   25.68 |  2.21e+03 (x 1.0) |  3.19e+03 (x 1.0) |  3.31e+03 |  3.31e+03 |
| CPP/sse4   |   24.72 =  11.49 +   13.24 |  3.31e+03 (x 1.5) |  6.19e+03 (x 2.0) |  6.55e+03 |  6.54e+03 |
| CPP/avx2   |   16.80 =  11.48 +    5.32 |  4.88e+03 (x 2.2) |  1.54e+04 (x 4.9) |  1.60e+04 |  1.59e+04 |
| CPP/512y   |   16.46 =  11.48 +    4.99 |  4.98e+03 (x 2.3) |  1.64e+04 (x 5.2) |  1.70e+04 |  1.66e+04 |
| CPP/512z   |   14.44 =  11.45 +    2.99 |  5.67e+03 (x 2.6) |  2.74e+04 (x 8.7) |  2.81e+04 |  2.82e+04 |
===========================================================================================================


+++  REVISION 667b080 (commit date: 2024-10-12 12:56:31 +0200) +++
On itscrd90.cern.ch [CPU: Intel(R) Xeon(R) Silver 4216 CPU] [GPU: 1x Tesla V100S-PCIE-32GB]:

===========================================================================================================
|            | mad                        | mad               | mad               | sa/brdg   | sa/full   |
-----------------------------------------------------------------------------------------------------------
| guxtaptamggux  | [sec] tot = mad + MEs  | [TOT/sec]         | [MEs/sec]         | [MEs/sec] | [MEs/sec] |
===========================================================================================================
| nevt/grid  |                            |                   |                   |      8192 |      8192 |
| nevt total |                      81920 |             81920 |             81920 |  256*32*1 |  256*32*1 |
-----------------------------------------------------------------------------------------------------------
| FORTRAN    |   52.01 =  16.88 +   35.13 |  1.58e+03 (= 1.0) |  2.33e+03 (= 1.0) |       --- |       --- |
| CPP/none   |   50.86 =  16.92 +   33.94 |  1.61e+03 (x 1.0) |  2.41e+03 (x 1.0) |  2.52e+03 |  2.52e+03 |
| CPP/sse4   |   33.94 =  16.93 +   17.01 |  2.41e+03 (x 1.5) |  4.82e+03 (x 2.1) |  5.00e+03 |  5.01e+03 |
| CPP/avx2   |   24.78 =  17.18 +    7.59 |  3.31e+03 (x 2.1) |  1.08e+04 (x 4.6) |  1.13e+04 |  1.12e+04 |
| CPP/512y   |   24.06 =  17.10 +    6.96 |  3.40e+03 (x 2.2) |  1.18e+04 (x 5.1) |  1.24e+04 |  1.24e+04 |
| CPP/512z   |   26.51 =  16.95 +    9.56 |  3.09e+03 (x 2.0) |  8.57e+03 (x 3.7) |  8.56e+03 |  8.71e+03 |
| CUDA/8192  |   17.66 =  17.40 +    0.25 |  4.64e+03 (x 2.9) |  3.23e+05 (x138.) |  2.75e+05 |  3.06e+05 |
===========================================================================================================
| nevt/grid  |                                                                    |     16384 |     16384 |
| nevt total |                                                                    |  512*32*1 |  512*32*1 |
--------------                                                                    -------------------------
| CUDA/max   |                                                                    |  4.32e+05 |  4.58e+05 |
|            |                                                                    |           |   (x196.) |
==============                                                                    =========================

… fix this so that only CUDA or HIP is printed out
…ut - still need to exclude GPU tags in CPU-only tests
…ee separate chep24 tables for rd90/gold/lumi
…to the subprocess of pp_dy3j I focused on in the cmsdy branch)

Note: there is no need to use no_b_mass to test phase space sampling in this specific process
CUDACPP_RUNTEST_DUMPEVENTS=1 ./build.cuda_m_inl0_hrd0/runTest_cuda.exe
\cp ../../test/ref/dump* ../../../CODEGEN/PLUGIN/CUDACPP_SA_OUTPUT/test/ref/
./tput/teeThroughputX.sh -guxtaptamggux -makej -makeclean
./tmad/teeMadX.sh -guxtaptamggux +10x
./tput/teeThroughputX.sh -guxtaptamggux -dmf -makej

STARTED AT Sat Oct 12 12:24:06 PM CEST 2024
ENDED   AT Sat Oct 12 12:25:19 PM CEST 2024
./tmad/teeMadX.sh -guxtaptamggux -dmf +10x

STARTED AT Sat Oct 12 12:33:31 PM CEST 2024
ENDED   AT Sat Oct 12 12:42:04 PM CEST 2024
./tput/teeThroughputX.sh -guxtaptamggux -dmf -makej

STARTED AT Sat Oct 12 12:25:24 PM CEST 2024
ENDED   AT Sat Oct 12 12:29:17 PM CEST 2024
./tmad/teeMadX.sh -guxtaptamggux -dmf +10x

STARTED AT Sat Oct 12 12:33:25 PM CEST 2024
ENDED   AT Sat Oct 12 12:47:14 PM CEST 2024
…o describe itscrd90/itgold91 d/m/f tests of guxtaptamggux
…build summaryTable_dy3j_chep24.txt for guxtaptamggux and add it to the repo
@valassi valassi self-assigned this Oct 12, 2024
@oliviermattelaer
Copy link
Member

I would say that this is the example of stuff that we agreed to remove from the repo, so this should NOT be included.
(Those results should be put in an overleaf or to another repo but not in this repo.

(In top of that 50 commits is also making the history even worse than what it is and I know that you will not accept to squash those)

Cheers,

Olivier

@valassi
Copy link
Member Author

valassi commented Oct 12, 2024

Hi Olivier, thanks :-)

Ok no problem, I will cleanup after CHEP and remove all the stuff that already exists.

As for these new data and table ok not to merge them then. I will keep this kind of stuff in my private fork.

Andrea

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants