Name
Title | ||
|---|---|---|
Efficient heterogeneous execution on large multicore and accelerator platforms: Case study using a block tridiagonal solver |
Abstract | ||
|---|---|---|
The algorithmic and implementation principles are explored in gainfully exploiting GPU accelerators in conjunction with multicore processors on high-end systems with large numbers of compute nodes, and evaluated in an implementation of a scalable block tridiagonal solver. The accelerator of each compute node is exploited in combination with multicore processors of that node in performing block-level linear algebra operations in the overall, distributed solver algorithm. Optimizations incorporated include: (1) an efficient memory mapping and synchronization interface to minimize data movement, (2) multi-process sharing of the accelerator within a node to obtain balanced load with multicore processors, and (3) an automatic memory management system to efficiently utilize accelerator memory when sub-matrices spill over the limits of device memory. Results are reported from our novel implementation that uses MAGMA and CUBLAS accelerator software systems simultaneously with ACML (2013) [2] for multithreaded execution on processors. Overall, using 940 nVidia Tesla X2090 accelerators and 15,040 cores, the best heterogeneous execution delivers a 10.9-fold reduction in run time relative to an already efficient parallel multicore-only baseline implementation that is highly optimized with intra-node and inter-node concurrency and computation-communication overlap. Detailed quantitative results are presented to explain all critical runtime components contributing to hybrid performance. |
Year | DOI | Venue |
|---|---|---|
2013 | 10.1016/j.jpdc.2013.07.012 | J. Parallel Distrib. Comput. |
Keywords | Field | DocType |
device memory,block tridiagonal solver,multicore processor,accelerator platform,accelerator memory,large multicore,gpu accelerator,case study,efficient heterogeneous execution,x2090 accelerator,novel implementation,automatic memory management system,cublas accelerator software system,implementation principle,efficient memory mapping,memory management,linear algebra | Synchronization,Concurrency,CUDA,Computer science,Parallel computing,Software system,Memory management,Solver,Multi-core processor,Distributed computing,Scalability | Journal |
Volume | Issue | ISSN |
73 | 12 | 0743-7315 |
Citations | PageRank | References |
3 | 0.40 | 16 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Alfred J. Park | 1 | 35 | 4.53 |
| Kalyan S. Perumalla | 2 | 725 | 64.15 |