CMS Mg5amc@NLO integration

Tuesday 10 Oct 2023, 15:00 → 16:00 Europe/Zurich

513/R-070 - Openlab Space (CERN)

# Tue 10.10.2023 Madgraph meeting with CMS
https://indico.cern.ch/event/1333115/

Present: SR, AV, Sapta, Jin, Robert, Zenny

## DY+4jets

Sapta: tried DY+3 jets is ok, but DY+4 jets failed compilation
SR: general problem, 6 final state particles fails
Jin: details in hedgedoc https://demo.hedgedoc.org/SXPUzCZqQWyDWkC9paHbVA#

AV: fails in generation (output) or build (launch)?
Jin: it fails in build/launch

SR: what is the goal of this?
Sapta: would like to put DY+4jets in production

AV: see https://github.com/madgraph5/madgraph4gpu/pull/601 with tests I had done in ggttgggg (2>6)
There it seems that it was the generation of fortran that failed (while generating sa cpp/gpu was ok)
AV: does DY+4jets work in the fortran production version?
Sapta/Robert: we have some data sets, so normally this works

## Flamegraphs

Jin: was unable to run perf, does it need root access?
SR: will ask SH, not sure
AV: cannot remember, but check for instance https://unix.stackexchange.com/a/14256

## Priorities?

Sapta: for us it would be the DY+4jet and the visualization

## Workshop

Sapta: what is expected of experiments?
SR: experiments will give a talk
AV: for CMS it is Bugra, you should talk with him
Sapta: ok (Bugra is replacing Sapta as CMS generator convenor)

Sapta: but there is also a Madgraph/GPU talk?
SR: yes this will probably be us developers giving this... for CMS it is better you use the CMS talk slot

## NLO

ZW: doing some profiling

AV: meeting with Alessandro Vicini tomorrow, there might be something related
ZW: will try to be there

## AOB

SR: should we have regular meetings biweekly?
Sapta: this slot works for me
SR: tentatively lets meet 24 October, otherwise we'll cancel

 

 

 

15:00 → 15:50 Discussion

Speakers: Jin Choi (Seoul National University (KR)), Saptaparna Bhattacharya (Northwestern University (US))

0040-fix-issue-use_syst-gpu.patch

Hedgedoc with progress

ME5_debug