Both Belle II and HyperK Data Acquisition requirements are very similar to those of other leading experiments at the intensity frontier. In particular the search for rare processes require very high trigger rates.
Belle II data taking is designed for trigger rates of 30 kHz, which is a factor >50 larger than in the former Belle experiment. The High Level Trigger acceptance rate is planned to be 10 kHz. A DAQ upgrade at Belle II is now being discussed, with the aim of increasing the HLT accept rate to open up the window for possible detection of new physics.
The HyperK DAQ and trigger system, presently under construction using ToolDAQ, has single rates of photomultiplier tube (PMT) dark noise of more than 10 kHz for each of the 40,000 PMTs. The use of sophisticated algorithms will lower this to trigger rate of between ~10-40kHz depending on the algorithm, where low energy running (e.g. solar neutrinos, supernovae), will dominate these random coincidences.
These systems and possible upgrades, require:
(i) new hardware technologies for high bandwidth data transfer, in particular optical technologies up to 16.3 Gbps on newest generation FPGAs, or 10 Gbps ethernet.
(ii) precise timing distribution in the sub-nanosecond regime. Possibly the inclusion of White Rabbit.
(iii) intelligent realtime algorithms for online data reduction. Belle II uses online tracking on an FPGA for background rejection. Background means on the one hand beam background and physics background e.g. from QED processes. HyperK plans to multiple trigger algorithms including vertexing on trigger level (t < 10 ns) on GPUs and large data buffers for supernovas.
(iv) novel programming and DAQ software techniques, such as methods of parallelisation on both FPGAs or GPUs, methods of artificial intelligence for trigger decisions with multiple input parameters, integrated dynamic service discovery, monitoring, fault tolerance, dynamic routing and remote control.
We foresee an active knowledge exchange between the two experiments, in form of combined DAQ technology and trigger workshops, which will be organized either jointly with annual JENNIFER2 general meetings, either in collaboration with Belle II or HyperK DAQ workshops, and at least once in an independent way, during the second half of the project, involving also private sector technology partners.
