present: Enrico, Roland, Milo, Frank, Andrea, Dan Gastler
Andrea:
- had success with the intended setup (sending data in Full mode from one VC709 card to another, via the optical link)
- however, the FELIX side is still not stable (e.g., probably the issue with incorrect chunk types is not yet fixed), so a continuous data transfer does not work yet
- need to see how quickly this can be done; but it should help that Nikhef, in the context of the FELIX project, now has a bit more time to devote to Full mode
- next steps:
- want to make the full mode generator more generic (so as to be useful to ATLAS as well)
- also want to enable a local loop-back test (so that a second FELIX card will not be needed)
- pointed out that operation in GBT mode could be useful also to our team on ProtoDUNE (even if not for ProtoDUNE operation itself)
Conclusion: we should go ahead with our plans for MSW3 at CERN next week
- hardware: Enrico and Roland will ask Giovanna for a second VC709 card (presumably a second PC will not be necessary)
Dan: WIB firmware
- received firmware block, but PCS (part of 8b/10b encoding) does not work natively on Altera above 6 Gb/s (and we need 9.6 Gb/s)
- so rather than using the provided functionality, re-coded this as part of the firmware "fabric" (also for RCE, for simplicity)
- can make this available to ATLAS as well (this would be really useful, according to Andrea)
- using 64-bit data internally (not 32 bits as for FELIX), so not easy to use identical transceiver wrappers
- WIB hardware: expect a board from BNL any day. Same hardware for FELIX as for RCE, just firmware differences
- can test FELIX transmission (at BU) once a new BNL 711 board becomes available (but according to Andrea this will not be before Q4 2017)
- in principle could test already using AREA5 (sp?) functionality on SBND WIB (but will need GP version to exceed 5 Gb/s)
We will meet next on May 17, to discuss the outcome of the MSW3 and the path forward.
There are minutes attached to this event.
Show them.