BEGIN:VCALENDAR
PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN
VERSION:1.0
BEGIN:VEVENT
DTSTART:20111115T193000Z
DTEND:20111115T200000Z
LOCATION:TCC 305
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: This paper introduces a novel implementation in reducing a symmetric dense matrix to tridiagonal form, which is the preprocessing step toward solving symmetric eigenvalue problems. Based on tile algorithms, the reduction follows a two-stage approach, where the tile matrix is first reduced to symmetric band form prior to the final condensed structure. The challenging trade-off between algorithmic performance and task granularity has been tackled through a grouping technique, which consists in aggregating fine-grained and memory-aware computational tasks during both stages, while sustaining the application overall high performance. A dynamic runtime environment system schedules then the different tasks in an out-of-order fashion.=0A=0AThe performance for the tridiagonal reduction reported in this paper =0Aare unprecedented. Our implementation results in an up to 50-fold=0Aimprovement (125 Gflop/s) compared to the equivalent routine from LAPACK=0Aand Intel MKL on an eight socket hexa-core AMD Opteron multicore shared-memory system with a matrix size of 24000x24000.
SUMMARY:Parallel Reduction to Condensed Forms for Symmetric Eigenvalue Problems using Aggregated Fine-Grained and Memory-Aware Kernels
PRIORITY:3
END:VEVENT
END:VCALENDAR