High-per­for­mance com­pu­ting (HPC) has become an important method in many sci­en­ti­fic fields, such as drug and cli­mate rese­arch. Howe­ver, the data cen­tres with energy-inten­sive super­com­pu­ters release quite a bit of CO2. “Energy-opti­mi­sed super­com­pu­ter net­works through the use of wind energy” (ESN4NW): Behind this title is a new, nati­on­wide joint pro­ject led by the SICP — Soft­ware Inno­va­tion Cam­pus Pader­born at the Uni­ver­sity of Pader­born. In coope­ra­tion with the SICP mem­ber com­pany West­fa­len­WIND IT and their brand wind­CORES, the idea was born to deve­lop a new HPC infra­struc­ture with a sus­tainable con­cept. The Fede­ral Minis­try of Edu­ca­tion and Rese­arch (BMBF) is now fun­ding the pro­ject for the next three years with around 2.5 mil­lion euros. The con­sor­tium also includes the Uni­ver­sity of Pas­sau, the Fraun­ho­fer Insti­tute for Relia­bi­lity and Micro­in­te­gra­tion IZM and the com­pa­nies Aix­pert­Soft GmbH, Rit­tal, Atos Deutsch­land and Zattoo.

New infrastructure with data centres in wind turbines

.

Tog­e­ther, the part­ners are rese­ar­ching and demons­t­ra­ting the poten­tial of so-cal­led “wind­CORES”. These are wind tur­bi­nes that house data cen­tres in their towers and sup­ply them with the wind power gene­ra­ted on site in an almost cli­mate-neu­tral man­ner. The aim of the pro­ject is to deve­lop the infra­struc­ture and ope­ra­tio­nal manage­ment of an HPC clus­ter within seve­ral wind tur­bi­nes (WTs). The direct, locally available rene­wa­ble energy is to be incor­po­ra­ted into the ope­ra­tio­nal manage­ment in order to make maxi­mum use of it. In addi­tion, the waste heat gene­ra­ted is also to be con­side­red as a limi­ting factor.

For demons­tra­tion pur­po­ses, an alter­na­tive — bey­ond the state of the art — archi­tec­tu­ral approach will be pur­sued and imple­men­ted. The pro­ject part­ners are deve­lo­ping an infra­struc­ture that con­ti­nuously adapts the com­pu­ting power of the decen­tra­li­sed HPC clus­ter through inten­sive data uti­li­sa­tion and arti­fi­cial intel­li­gence (AI) methods via an ope­ra­ting plat­form. The clus­ter is dis­tri­bu­ted over seve­ral WTs and must opti­mally exploit the fluc­tua­ting energy avai­la­bi­lity. The pro­ject also explo­res the impact of dif­fe­rent archi­tec­tures and ope­ra­ting stra­te­gies on the sus­taina­bi­lity assess­ment of the tar­get sys­tem. AI-based ope­ra­tion manage­ment requi­res beha­viou­ral and pre­dic­tive models of the infra­struc­ture and the local energy grid, which are not yet available. “The sci­en­ti­fic and tech­ni­cal chal­lenges of the pro­ject lie, among other things, in the inten­sive data deve­lo­p­ment and use (energy grid & infra­struc­ture), in the inte­gra­tion of arti­fi­cial intel­li­gence into con­trol loops, as well as in the sys­te­mic mer­ging of all models with the help of a digi­tal twin,” explains Dr. Gun­nar Scho­ma­ker, mana­ger at SICP. “The focus is on pre­dic­tions of wea­ther-depen­dent local energy avai­la­bi­lity, as well as pre­cise beha­viou­ral models of all inter­ac­ting trades with regard to their power con­sump­tion and heat out­put,” Scho­ma­ker adds.

Innovative concept with intelligent circuitry

A spe­cial fea­ture and inno­va­tion of the pro­ject is that the tower of a wind tur­bine is inte­gra­ted as a poten­tial heat sink if no waste heat recei­ver is available or if it does not need the waste heat at the moment. The tower can then absorb the waste heat from the IT sys­tems and cool it down. This func­tion, tog­e­ther with energy avai­la­bi­lity, is to be used as a gui­de­line for the ope­ra­tion of the HPC clus­ter. In con­crete terms, this means that the sys­tem will only run if rene­wa­ble energy is available and waste heat can be dis­si­pa­ted in an almost CO2-neu­tral man­ner. Initi­ally, the pro­ject will focus on the tech­ni­cal and eco­no­mic fea­si­bi­lity, the func­tional and didac­tic demons­tra­tion of the ope­ra­tion and the sus­taina­bi­lity assess­ment of the sys­tem in the exten­ded sys­tem life cycle. Howe­ver, the inten­ded results should be trans­fera­ble not only to wind tur­bi­nes, but also to other HPC sys­tems with waste heat gene­ra­tors and waste heat users.

“The pro­ject par­ti­cu­larly addres­ses the chal­lenges of the energy tran­si­tion and digi­ta­li­sa­tion. We want to show that the incre­asing energy demand of digi­ta­li­sa­tion is not a dead end for more sus­taina­bi­lity and that these growth needs can also be covered fle­xi­bly in terms of time and space by rege­ne­ra­tive ener­gies,” says Dr Fiete Dub­berke, Mana­ging Direc­tor of West­fa­len­WIND IT, describ­ing the pro­ject goals. “In doing so, we are asking our­sel­ves the fol­lo­wing ques­ti­ons: How can we deal with waste heat from the most diverse indus­trial sec­tors, i.e. use it? Which sys­tems of gene­ra­tors and users are to be con­nec­ted in a meaningful way and what tasks can arti­fi­cial intel­li­gence methods take over?”, Scho­ma­ker ela­bo­ra­tes. The “wind­CORES” con­cept, which was awarded the “Ger­man Data Centre Prize”, has alre­ady shown that mer­ging a wind tur­bine with a data centre can also be imple­men­ted suc­cessfully in eco­no­mic terms.