The success of the throughput-optimised (ThOr) functors in the x86 HLT1 raises the question of whether we can profit from them in HLT2.

Recent studies have shown very promising results with an example D^{+} \to h^{+}h^{-}h^{+} HLT2 line. Further gains have been shown when adding a new, experimental event model.

Transitioning to a brand new event model is a huge undertaking, requiring support of a wide set of physics use-cases and the rewrite of many tools and services. Adding support for the existing LHCb::Particle-based model to ThOr functors is however a more easily tractable task.

The aim of this milestone is to expand the set of available functors and associated algorithms such that a suite of HLT2 lines can utilise ThOr functors. This will allow us to run more detailed benchmarks with a wider set of realistic physics selections.

Ideally, work on the (possible) future event model will continue in parallel, but this is subject to finding new interested people.