Automatic Parallelization Techniques Based on Compact Dag Extraction and Symbolic Scheduling

Cosnard, Michel; Jeannot, Emmanuel

doi:10.1142/s012962640100049x

Cited by 8 publications

(5 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…3The third approach consists of using the concise representation of the DAG in memory, to avoid most of the impact of unrolling it at runtime. Using structures like Parameterized Task Graph (PTG) [36], the memory used for DAG representation is linear in the number of task types and totally independent of the total number of tasks.…”

Section: Runtime System: Smpss/quarkmentioning

confidence: 99%

Two-Stage Tridiagonal Reduction for Dense Symmetric Matrices Using Tile Algorithms on Multicore Architectures

Łuszczek

Ltaief

Dongarra

2011

2011 IEEE International Parallel &Amp; Distributed Processing Symposium

View full text Add to dashboard Cite

While successful implementations have already been written for one-sided transformations (e.g., QR, LU and Cholesky factorizations) on multicore architecture, getting high performance for two-sided reductions (e.g., Hessenberg, tridiagonal and bidiagonal reductions) is still an open and difficult research problem due to expensive memorybound operations occurring during the panel factorization. The processor-memory speed gap continues to widen, which has even further exacerbated the problem. This paper focuses on an efficient implementation of the tridiagonal reduction, which is the first algorithmic step toward computing the spectral decomposition of a dense symmetric matrix. The original matrix is translated into a tile layout i.e., a high performance data representation, which substantially enhances data locality. Following a twostage approach, the tile matrix is then transformed into band tridiagonal form using compute intensive kernels. The band form is further reduced to the required tridiagonal form using a left-looking bulge chasing technique to reduce memory traffic and memory contention. A dependence translation layer associated with a dynamic runtime system allows for scheduling and overlapping tasks generated from both stages. The obtained tile tridiagonal reduction significantly outperforms the state-of-theart numerical libraries (10X against multithreaded LAPACK with optimized MKL BLAS and 2.5X against the commercial numerical software Intel MKL) from medium to large matrix sizes.

show abstract

Section: Runtime System: Smpss/quarkmentioning

confidence: 99%

Two-Stage Tridiagonal Reduction for Dense Symmetric Matrices Using Tile Algorithms on Multicore Architectures

Łuszczek

Ltaief

Dongarra

2011

2011 IEEE International Parallel &Amp; Distributed Processing Symposium

View full text Add to dashboard Cite

show abstract

“…This was the starting point to several paper related to work stealing scheduler optimization which is outside the scope of our TDG optimization. Parsec [33] based their graph model on the parametrized graph model [34] to implicitly represent tasks and their dependencies. It was a huge step in optimizing the memory space related to a task graph.…”

Section: Related Workmentioning

confidence: 99%

Investigating Dependency Graph Discovery Impact on Task-based MPI+OpenMP Applications Performances

Pereira

Roussel

Carribault

et al. 2023

Proceedings of the 52nd International Conference on Parallel Processing

View full text Add to dashboard Cite

The architecture of supercomputers is evolving to expose massive parallelism. MPI and OpenMP are widely used in application codes on the largest supercomputers in the world. The community primarily focused on composing MPI with OpenMP before its version 3.0 introduced task-based programming. Recent advances in OpenMP task model and its interoperability with MPI enabled fine model composition and seamless support for asynchrony. Yet, OpenMP tasking overheads limit the gain of task-based applications over their historical loop parallelization (parallel for construct).This paper identifies the OpenMP task dependency graph discovery speed as a limiting factor in the performance of task-based applications. We study its impact on intra and inter-node performances over two benchmarks (Cholesky, HPCG) and a proxyapplication (LULESH). We evaluate the performance impacts of several discovery optimizations, and introduce a persistent task dependency graph reducing overheads by a factor up to 15 at run-time. We measure 2x speedup over parallel for versions weak scaled to 16K cores, due to improved cache memory use and communication overlap, enabled by task refinement and depth-first scheduling.

show abstract

“…However, no consensus has still been reached on a specific paradigm. For example, Parametrized Task Graph approaches [Bosilca et al 2012;Budimlić et al 2010;Cosnard and Jeannot 2001] consist of explicitly describing tasks (vertices of the DAG) and their mutual dependencies (edges) by informing the runtime system with a set of dependency rules. In such a way, the DAG is never explicitly built but can be progressively unrolled and traversed in a very effective and flexible way.…”

Section: Task-based Runtime Systems and The Sequential Task Flow Modelmentioning

confidence: 99%

Implementing Multifrontal Sparse Solvers for Multicore Architectures with Sequential Task Flow Runtime Systems

Agullo

Buttari

Guermouche

et al. 2016

ACM Trans. Math. Softw.

View full text Add to dashboard Cite

To face the advent of multicore processors and the ever increasing complexity of hardware architectures, programming models based on DAG parallelism regained popularity in the high performance, scientific computing community. Modern runtime systems offer a programming interface that complies with this paradigm and powerful engines for scheduling the tasks into which the application is decomposed. These tools have already proved their effectiveness on a number of dense linear algebra applications. This paper evaluates the usability and effectiveness of runtime systems based on the Sequential Task Flow model for complex applications, namely, sparse matrix multifrontal factorizations which feature extremely irregular workloads, with tasks of different granularities and characteristics and with a variable memory consumption. Most importantly, it shows how this parallel programming model eases the development of complex features that benefit the performance of sparse, direct solvers as well as their memory consumption. We illustrate our discussion with the multifrontal QR factorization running on top of the StarPU runtime system.

show abstract

Automatic Parallelization Techniques Based on Compact Dag Extraction and Symbolic Scheduling

Cited by 8 publications

References 18 publications

Two-Stage Tridiagonal Reduction for Dense Symmetric Matrices Using Tile Algorithms on Multicore Architectures

Two-Stage Tridiagonal Reduction for Dense Symmetric Matrices Using Tile Algorithms on Multicore Architectures

Investigating Dependency Graph Discovery Impact on Task-based MPI+OpenMP Applications Performances

Implementing Multifrontal Sparse Solvers for Multicore Architectures with Sequential Task Flow Runtime Systems

Contact Info

Product

Resources

About