1 / 4

Multiple Component Multiple Data Mini Workshop

Multiple Component Multiple Data Mini Workshop. Salish Lodge January 24, 2007. Multilevel Parallelism. How can applications effectively exploit the massive amount of parallelism available in teraflop and future petaflop-scale machines?

dreama
Download Presentation

Multiple Component Multiple Data Mini Workshop

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Multiple Component Multiple DataMini Workshop Salish Lodge January 24, 2007

  2. Multilevel Parallelism • How can applications effectively exploit the massive amount of parallelism available in teraflop and future petaflop-scale machines? • Massive numbers of CPUs in future systems require algorithm and software redesidgn to exploit all available parallelism • A proven solution is • To exploit fully available hardware parallelism for applications, exploitation of multiple levels of parallelism (MLP) • Hierarchical parallelism – algorithm decomposition at different levels • Increases granularity of computation => improve the overall scalability.

  3. Multiple Component Multiple Data • MCMD extends the SCMD (single component multiple data) model that was the main focus of CCA in Scidac-1 • Prototype solution described at SC’05 for computational chemistry • Allows different groups of processors execute different CCA components • Main motivation is support for multiple levels of parallelism in applications

  4. Issues • Investigate requirements in existing and emerging CCA applications • Define extensions to the CCA specs? • Processor group management in CCA • Enhancements to the CCA core software stack (frameworks, services) • MCMD-awareness in the scientific component toolkit • Context of different programming models (MPI, PVM, GA, OpenMP, Pthreads, GAS languages, DARPA HPCS languages) • Relationship and collaboration with other CCA activities like CQoS, CCA on hybrid platforms, fault tolerance • External collaborations and synergistic activities elsewhere

More Related