The ALICE experiment has undergone a major upgrade for LHC Run 3 and will record 50 times more heavy ion collisions than before.
The new computing scheme for Run 3 replaces the traditionally separate online and offline frameworks by a unified one.
Processing will happen in two phases.
During data taking, a synchronous processing phase performs data compression, calibration, and quality control on the online computing farm.
The output is stored on an on-site disk buffer.
When there is no beam in the LHC, the same computing farm is used for the asynchronous reprocessing of the data which yields the final reconstruction output.
ALICE will employ neither hardware nor software triggers for Pb-Pb data taking but instead store all collisions in compressed form.
This requires full online processing of all recorded data, which is a major change compared to a traditional online systems, which sees only the data selected by a hardware trigger.
Traditional CPUs are unable to cope with the huge data rate and processing demands of the synchronous phase, therefore ALICE employs GPUs to speed up the processing.
Since the online computing farm performs a part of the asynchronous processing, ALICE plans to use the GPUs also for this second phase when there is no beam in the LHC.
The primary goal for the commissioning in 2021 and 2022 was to make those reconstruction steps required for the online phase run on the GPU efficiently.
The development is now shifting towards moving more computing-intensive steps of the asynchronous reconstruciton to the GPU as well.
The talk will detail the ALICE Run 3 computing scheme, and outline the hardware architecture and software design for synchronous and asynchronous processing.