Accelerating End-to-End Deep Learning for Particle Reconstruction using CMS open data

Machine learning algorithms are gaining ground in high energy physics for applications particle and event identification, analysis, detector reconstruction, simulation trigger. Currently, most data-analysis tasks at LHC experiments benefit from the use of machine learning. Incorporating these computational tools experimental framework presents new challenges. This paper reports on implementation end-to-end deep with CMS software scaling multiple GPUs. The technique combines low-level representation identification. We demonstrate a top quark benchmark perform studies various hardware architectures including single GPUs Google TPU.