Skip to content

Commit

Permalink
[gqttq] ** COMPLETE GQTTQ** revert tmad test and log for gqttq from P…
Browse files Browse the repository at this point in the history 
…1_gux_ttxux to P1_gu_ttxu

The gqttq tests fail anyway and will need to be fixed (madgraph5#630).
However, this completes the addition of gq_ttq as a new process to the repo.
In particular it includes proof that Olivier's "split_nonidentical_grouping" madgraph5#619 fixes the gqttq builds.
It also includes a lot of cleanup for "nprocesses" (madgraph5#272 and madgraph5#343)

Revert "[gqttq] retry the tmad gqttq test with the P1_gu_ttxu directory - the test continues to fail (madgraph5#630)"
This reverts commit 2dea1f7.

Revert "[gqttq] temporarely use P1_gu_ttxu instead of P1_gux_ttxux for gqttq tmad tests"
This reverts commit ea23a9a.
  • Loading branch information
@valassi
valassi committed Apr 5, 2023
1 parent 2dea1f7 commit 113fef5
Show file tree
Hide file tree
Showing 2 changed files with 39 additions and 39 deletions.
74 changes: 37 additions & 37 deletions epochX/cudacpp/tmad/logs_gqttq_mad/log_gqttq_mad_d_inl0_hrd0.txt
View file
Original file line number Diff line number Diff line change
@@ -1,42 +1,42 @@
Working directory (build): /data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gu_ttxu
Working directory (build): /data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gux_ttxux
CUDACPP_BUILDDIR='.'



make USEBUILDDIR=1 AVX=none
make USEBUILDDIR=1 AVX=sse4
make USEBUILDDIR=1 AVX=avx2

make USEBUILDDIR=1 AVX=avx2
make USEBUILDDIR=1 AVX=512y
make[1]: Entering directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gu_ttxu'
make[1]: Entering directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gu_ttxu'
make[1]: Entering directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gu_ttxu'
make[1]: Entering directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gu_ttxu'
make[1]: Entering directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gux_ttxux'
make[1]: Entering directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gux_ttxux'
make[1]: Entering directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gux_ttxux'
make[1]: Entering directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gux_ttxux'

make USEBUILDDIR=1 AVX=512z
make[1]: Entering directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gu_ttxu'
make[1]: Entering directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gux_ttxux'
CUDACPP_BUILDDIR='build.none_d_inl0_hrd0'
make[1]: Nothing to be done for 'all'.
make[1]: Leaving directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gu_ttxu'
CUDACPP_BUILDDIR='build.avx2_d_inl0_hrd0'
CUDACPP_BUILDDIR='build.512y_d_inl0_hrd0'
CUDACPP_BUILDDIR='build.512z_d_inl0_hrd0'
make[1]: Nothing to be done for 'all'.
make[1]: Leaving directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gu_ttxu'
CUDACPP_BUILDDIR='build.sse4_d_inl0_hrd0'
make[1]: Leaving directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gux_ttxux'
make[1]: Nothing to be done for 'all'.
make[1]: Leaving directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gu_ttxu'
make[1]: Leaving directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gux_ttxux'
make[1]: Nothing to be done for 'all'.
make[1]: Leaving directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gu_ttxu'
CUDACPP_BUILDDIR='build.512y_d_inl0_hrd0'
make[1]: Leaving directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gux_ttxux'
CUDACPP_BUILDDIR='build.sse4_d_inl0_hrd0'
make[1]: Nothing to be done for 'all'.
make[1]: Leaving directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gux_ttxux'
make[1]: Nothing to be done for 'all'.
make[1]: Leaving directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gu_ttxu'
make[1]: Leaving directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gux_ttxux'

OMP_NUM_THREADS=

DATE: 2023-04-05_12:04:45
DATE: 2023-04-05_11:53:39

On itscrd80.cern.ch [CPU: Intel(R) Xeon(R) Silver 4216 CPU] [GPU: 1x Tesla V100S-PCIE-32GB]:
Working directory (run): /data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gu_ttxu
Working directory (run): /data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gq_ttq.mad/SubProcesses/P1_gux_ttxux

*** (1) EXECUTE MADEVENT (create results.dat) ***
--------------------
Expand All @@ -56,10 +56,10 @@ Executing ' ./madevent < /tmp/avalassi/input_gqttq_x1_fortran > /tmp/avalassi/ou
[XSECTION] MultiChannel = TRUE
[XSECTION] Configuration = 1
[XSECTION] ChannelId = 1
[XSECTION] Cross section = 0.2605 [] fbridge_mode=0
[COUNTERS] PROGRAM TOTAL : 0.3682s
[COUNTERS] Fortran Overhead ( 0 ) : 0.2831s
[COUNTERS] Fortran MEs ( 1 ) : 0.0851s for 8192 events => throughput is 9.63E+04 events/s
[XSECTION] Cross section = 1.678 [] fbridge_mode=0
[COUNTERS] PROGRAM TOTAL : 0.3796s
[COUNTERS] Fortran Overhead ( 0 ) : 0.2941s
[COUNTERS] Fortran MEs ( 1 ) : 0.0855s for 8192 events => throughput is 9.58E+04 events/s

*** (1) EXECUTE MADEVENT x1 (create events.lhe) ***
--------------------
Expand All @@ -79,11 +79,11 @@ Executing ' ./madevent < /tmp/avalassi/input_gqttq_x1_fortran > /tmp/avalassi/ou
[XSECTION] MultiChannel = TRUE
[XSECTION] Configuration = 1
[XSECTION] ChannelId = 1
[XSECTION] Cross section = 0.2605 [0.26050333309703716] fbridge_mode=0
[UNWEIGHT] Wrote 81 events (found 765 events)
[COUNTERS] PROGRAM TOTAL : 0.3318s
[COUNTERS] Fortran Overhead ( 0 ) : 0.2484s
[COUNTERS] Fortran MEs ( 1 ) : 0.0834s for 8192 events => throughput is 9.82E+04 events/s
[XSECTION] Cross section = 1.678 [1.6775080845293466] fbridge_mode=0
[UNWEIGHT] Wrote 1611 events (found 1616 events)
[COUNTERS] PROGRAM TOTAL : 0.4877s
[COUNTERS] Fortran Overhead ( 0 ) : 0.4020s
[COUNTERS] Fortran MEs ( 1 ) : 0.0856s for 8192 events => throughput is 9.57E+04 events/s

*** (1) EXECUTE MADEVENT x10 (create events.lhe) ***
--------------------
Expand All @@ -103,11 +103,11 @@ Executing ' ./madevent < /tmp/avalassi/input_gqttq_x10_fortran > /tmp/avalassi/o
[XSECTION] MultiChannel = TRUE
[XSECTION] Configuration = 1
[XSECTION] ChannelId = 1
[XSECTION] Cross section = 0.218 [0.21801182648615874] fbridge_mode=0
[UNWEIGHT] Wrote 853 events (found 1849 events)
[COUNTERS] PROGRAM TOTAL : 2.3602s
[COUNTERS] Fortran Overhead ( 0 ) : 1.4365s
[COUNTERS] Fortran MEs ( 1 ) : 0.9237s for 90112 events => throughput is 9.76E+04 events/s
[XSECTION] Cross section = 0.8972 [0.89719199651173098] fbridge_mode=0
[UNWEIGHT] Wrote 1781 events (found 1786 events)
[COUNTERS] PROGRAM TOTAL : 2.4707s
[COUNTERS] Fortran Overhead ( 0 ) : 1.5497s
[COUNTERS] Fortran MEs ( 1 ) : 0.9209s for 90112 events => throughput is 9.78E+04 events/s

*** (2-none) EXECUTE CMADEVENT_CUDACPP x1 (create events.lhe) ***
--------------------
Expand All @@ -127,12 +127,12 @@ Executing ' ./build.none_d_inl0_hrd0/cmadevent_cudacpp < /tmp/avalassi/input_gqt
[XSECTION] MultiChannel = TRUE
[XSECTION] Configuration = 1
[XSECTION] ChannelId = 1
[XSECTION] Cross section = 1.276 [1.2757941949814184] fbridge_mode=1
[UNWEIGHT] Wrote 112 events (found 1776 events)
[COUNTERS] PROGRAM TOTAL : 0.5097s
[COUNTERS] Fortran Overhead ( 0 ) : 0.4393s
[COUNTERS] CudaCpp MEs ( 2 ) : 0.0705s for 8192 events => throughput is 1.16E+05 events/s
[XSECTION] Cross section = 4.263 [4.2628778044075899] fbridge_mode=1
[UNWEIGHT] Wrote 1344 events (found 3238 events)
[COUNTERS] PROGRAM TOTAL : 0.7038s
[COUNTERS] Fortran Overhead ( 0 ) : 0.6326s
[COUNTERS] CudaCpp MEs ( 2 ) : 0.0712s for 8192 events => throughput is 1.15E+05 events/s

*** (2-none) Compare CMADEVENT_CUDACPP x1 xsec to MADEVENT xsec ***

ERROR! xsec from fortran (0.26050333309703716) and cpp (1.2757941949814184) differ by more than 2E-14 (3.8974198518457603)
ERROR! xsec from fortran (1.6775080845293466) and cpp (4.2628778044075899) differ by more than 2E-14 (1.5411965782588837)
4 changes: 2 additions & 2 deletions epochX/cudacpp/tmad/madX.sh
View file
Original file line number Diff line number Diff line change
Expand Up @@ -137,8 +137,8 @@ function showdir()
elif [ "${ggttggg}" == "1" ]; then
dir=$topdir/epochX/${bckend}/gg_ttggg${suff}SubProcesses/P1_gg_ttxggg
elif [ "${gqttq}" == "1" ]; then
dir=$topdir/epochX/${bckend}/gq_ttq${suff}SubProcesses/P1_gu_ttxu
###dir=$topdir/epochX/${bckend}/gq_ttq${suff}SubProcesses/P1_gux_ttxux # only 1 out of 2 for now
###dir=$topdir/epochX/${bckend}/gq_ttq${suff}SubProcesses/P1_gu_ttxu
dir=$topdir/epochX/${bckend}/gq_ttq${suff}SubProcesses/P1_gux_ttxux # only 1 out of 2 for now
fi
else
echo "INTERNAL ERROR! tmad tests only make sense in .mad directories"; exit 1 # this should never happen (suff=.mad/ is hardcoded)
Expand Down

0 comments on commit 113fef5

Please # to comment.