Virtual process engineering (VPE) aims to redefine the roadmap for process scaling-up and optimization, from stepwise experiments to high-performance computer simulations. This is a long-cherished dream of chemical engineers, but requires high standards of Accuracy (the agreement between the simulation and the real process), Capability (the computational speed, scale, and resolution of the simulation), and Efficiency (cost-effective and easy to use), in short, ACE. For complex processes such as gas-solid fluidization, the gap between state-of-the-art simulations and VPE is still huge in terms of ACE. However, the work reported in this paper narrows this gap significantly. In this study, a coarse-grained discrete particle method (DPM) defined by the energy-minimization multi-scale (EMMS) model is deployed for high resolution simulations of fluidized beds, with the gas- and solid-phase equations solved concurrently by CPUs and GPUs in a heterogeneous supercomputing system. With systematic optimization of the model, numerical method, software, and hardware, we are able to simulate lab- to pilot-scale fluidized beds at quasi-realtime speed, and conduct virtual experiments on such systems. This enables very-long-time simulations to obtain important engineering parameters such as the particle residence time distribution, attrition and deactivation indexes. This work demonstrates that the industrial application of VPE is almost on the horizon. (C) 2016 Elsevier Ltd. All rights reserved.