b'Engineer Innovation | Aerospacebike in a wind tunnel model with about 12 million cells. Run time on an old, on-premise HPC device with 12 cores was six hours to extract one drag number. On the ScaleX cloud platform using 64 cores, this fell dramatically, to 42 minutes. Using 96 cores, run time was further compressed to 32 minutes. With 128 cores, run time was just 22 minutes. Instead of waiting almost a full workday for an answer, cloud HPC lets analysts extract an answer in close to real time, file the report and proceed to the next step in their project.Four-person drafting: identifying optimal rider assembly configuration for a given wind conditionIn a project demonstrating the value of cloud HPC, Trek studied a situation in cycling races where riders group closely together in whats called a pace line, to conserve energy by reducing wind drag. Using Simcenter STAR-CCM+ and HEEDS in a coupled analysis, engineers ran a simulation on ScaleX to study varying bicycle drafting methods. The case modeled four cyclists in the computational domain, each with X-Y and lateral coordinates. The objective was to identify the optimal rider assembly configuration for a given wind Figure 5: HEEDS output for water bottle placement optimization. condition. To begin, the CFD case was Figure 6: HEEDS output for the optimization of a four-person drafting configuration.32'