Project

General

Profile

Spring 2013 Collaboration Meeting

We mainly used the time to work, but we did have a discussion on the required trigger rates. Andrew provided the following numbers:

86 kB / Number of DCMs = data coming out of FEBs?
20 Mb/s from FEB (0 suppressed is much less)
=2.5 MB/s

86 kB for whole detector for 100 us (based on average event size: a bit of noise and some tracks) THIS IS NOT THE MAXIMUM EVENT SIZE!!! (just an average event size)
24 Mb/s DCM to Buffer Node (memory bandwitdth) = 4.3 Gb/s (=180 * 24 Mb/s) (10 Gb network, so it could handle ALL the traffic)

traffic into one buffer node = 1 Gb/s

buffer node to data logger = 2 Gb/s (2 1 Gb/s links)
in the future we might upgrade to 10 Gb/s

ash river to fermilab:
currently: 500 Mb/s
planned: 1 Gb/s
very far future: 10 Gb/s

300 TB disk at Far Detector
worst case scenario we have to carry disks from Ash River to Fermilab