
The finite element (FE)-code STRIPE was used during Capability Applications Project I (CAP-I) 2008 to efficiently and reliably solve ultra-large structures and materials problems having the order of a billion degrees of freedom on the IBM Power P6-system DaVinci. Problems of this size cannot be solved on the NAVY/Babbage system due to memory requirements or on the AFRL/Hawk system due to long execution wall-time requirements. STRIPE performance on the DaVinci and Babbage systems was compared and a factor of up to eight in speedup in wall-time was observed solving the largest problems solvable on Babbage. This significant increase in computational performance is due to a STRIPE code rewrite, a new finite element modeling (FEM)-approach, and the superior CPU and input/output (I/O) performance of the DaVinci system relative to Babbage. This paper describes the various techniques adopted during CAP-I on DaVinci to achieve high system scalability when solving, in short time, the world’s largest strength of materials problem related to aircraft maintenance and design. Support from major software vendors IBM, DoD Supercomputing Resource Center’s (DSRC’s) support specialists, as well as I/O specialists at ParaTools (as a part of the PET-program) have contributed to the successful CAP-I activities described here.
| selected citations These citations are derived from selected sources. This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | 1 | |
| popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network. | Average | |
| influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | Average | |
| impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network. | Average |
