MadGraph5_aMC@NLO

tt~bb~ process taking a long time to run.

Asked by Robert Cantrill

Hi

I’m trying to generate the p p > b b~ t t~ process at next-to-leading order. It is taking an extremely long time to run i.e. longer than a day for 10000 events. I’m guessing that at NLO there are a lot of diagrams to be generated and this is causing the long run time. From my output I see that MadGraph has generated 9 subprocesses with 1877 real emission diagrams, 92 born diagrams and 3730 virtual diagrams.

I was wondering if this is normal? And if not, is there anything I need to set in order to speed things up?

Kind regards,
Rob

Question information

Language:
English Edit question
Status:
Answered
For:
MadGraph5_aMC@NLO Edit question
Assignee:
Rikkert Frederix Edit question
Last query:
Last reply:
Revision history for this message
Rikkert Frederix (frederix) said : 		#1

Dear Rob,

I'm not at all surprised that a 2->4 process at NLO takes more than a day to run for 10k events. These calculations are extremely involved.

Note that the time does not scale exactly linear with the number of events. Per event, you'll probably get a better timing value if you generate more than 100k events in a single run.

Best regards,
Rikkert

Revision history for this message
Robert Cantrill (robert-cantrill) said : 		#2

Dear Rikkert

Thank you for your reply. I tried generating tth->bb events as well. This however runs a lot more quickly i.e ~1 hour. Is it surprising that this process would be a lot quicker?

Cheers,
Rob

Revision history for this message
Rikkert Frederix (frederix) said : 		#3

Dear Rob,

That's also not really a surprise. A 2->3 process with only 4 external QCD-charged particles is really a lot simpler than a 2->4 process with 6 QCD charged particles. You should already be able to see it from the number of Feynman diagrams.

Best regards,
Rikkert

Can you help with this problem?

Provide an answer of your own, or ask Robert Cantrill for more information if necessary.

To post a message you must log in.