CERN openlab: the use of GPUs for parallel computing within the LHCb experiment
As a member of CERN openlab, E4 has supported Allen, a key initiative to update CERN’s LHCb trigger.
Last week, the CERN LHC Experiments Committee formally accepted a proposal for a new phase of the high level trigger (HLT) for LHCb. LHCb is one of the four main experiments on the Large Hadron Collider (LHC), which studies what happened after the Big Bang and what phenomenon allowed matter to survive and build the universe we see today.
The new system used by CERN to do such experiment will be based on GPUs, which offer a highly parallelized structure and therefore greater performance when it comes to processing huge amounts of data in parallel.
Various algorithms have been efficiently implemented within the Allen project, demonstrating that GPUs have a very high potential, not only to be used as computational accelerators in the field of high energy physics, but also as complete and independent data processing solutions.
CERN has counted on the competence of our engineers with GPUs and on the strong link between our company and NVIDIA. As CERN openlab states, the link with companies like E4 is fundamental to obtain the best efficiency from the hardware. Read the article on CERN website.