WindHPC pur­sues the goal of making simu­la­tion-based pro­blems more sus­tainable. This is to be achie­ved by using eco­lo­gi­cal com­pu­ting resour­ces, incre­asing the effi­ci­ency of the soft­ware and opti­mi­zing work­flows. For this pur­pose, HPC resour­ces in wind tur­bi­nes are com­bi­ned with a super­com­pu­ting center. 

Our goals in this rese­arch project:

• The energy con­sump­tion of indi­vi­dual simu­la­ti­ons as well as of simu­la­tion work­flows is to be inves­ti­ga­ted on dif­fe­rent hard­ware and at dif­fe­rent loca­ti­ons, and for this pur­pose the deve­lo­p­ment of digi­tal twins is to be advanced
• Deve­lo­p­ment of soft­ware-side auto-tuning methods, which at run­time deter­mine the most energy-effi­ci­ent and, if neces­sary, the most sca­lable com­bi­na­tion of simu­la­tion resour­ces. sca­lable com­bi­na­tion of simu­la­tion algo­rithms, data struc­tures, and par­al­le­liza­tion strategies
• deve­lo­p­ment of intel­li­gent sche­du­ling methods that opti­mally dis­tri­bute and exe­cute the over­all work­flow across dif­fe­rent HPC clusters

, taking into account energy requi­re­ments and the tem­po­ral rele­vance of requi­red simulations.