Numerical simulations of Quantum Chromodynamics (QCD) are well established and a state-of-the-art method to obtain nonperturbative predictions from the Standard Model (SM) of elementary particle physics. These predictions are crucial to complement the experimental research program at the LHC or at the future Belle II experiment. In order to fully leverage the experimental results, uncertainties of the theoretical predictions need to be at most of the same size and thus have to keep track with ongoing experimental improvements, higher luminosities, and refined detectors. In order to further reduce uncertainties of lattice QCD simulations, algorithmic improvements are required and our simulations codes need to account for the changes in the design of the next generation of supercomputers. This workshop aims to bring together the leading experts in algorithms and hardware specific optimisation in order to discuss new paradigms and methods required for future machines as well as challenges of simulations with physical light and heavy quarks on larger and finer lattices. In addition we aim to facilitate collaborations among the participants with the intention to agree on common interfaces allowing for higher portability of code packages.