CilkSpec: Optimistic Concurrency for Cilk

Aga, Shaizeen D.; Krishnamoorthy, Sriram; Narayanasamy, Satish

doi:10.1145/2807591.2807597

Title: CilkSpec: Optimistic Concurrency for Cilk

Full Record
Other Related Research

Abstract

Recursive parallel programming models such as Cilk strive to simplify the task of parallel programming by enabling a simple divide-and-conquer model of programming. This model is effective in re- cursively partitioning work into smaller parts and combining their results. However, recursive work partitioning can impose additional constraints on concurrency than is implied by the true dependences in a program. In this paper, we present a speculation-based approach to alleviate the concurrency constraints imposed by such recursive parallel programs. We design a runtime infrastructure that supports speculative execution and a predictor to accurately learn and identify opportunities to relax extraneous concurrency constraints. Experimental evaluation demonstrates that speculative relaxation of concurrency constraints can deliver gains of up to 1.6× on 30 cores over baseline Cilk.

Authors:: Aga, Shaizeen D.; Krishnamoorthy, Sriram; Narayanasamy, Satish

Publication Date:: Sun Nov 15 00:00:00 EST 2015

Research Org.:: Pacific Northwest National Lab. (PNNL), Richland, WA (United States)

Sponsoring Org.:: USDOE

OSTI Identifier:: 1236312

Report Number(s):: PNNL-SA-111827
KJ0402000

DOE Contract Number:: AC05-76RL01830

Resource Type:: Conference

Resource Relation:: Conference: Supercomputing (SC15): Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, November 15-20, 2015, Austin, Texas, Paper No. 83

Country of Publication:: United States

Language:: English

Subject:: fork-join; cilk; speculation; optimistic concurrency

Citation Formats


                    Aga, Shaizeen D., Krishnamoorthy, Sriram, and Narayanasamy, Satish. CilkSpec: Optimistic Concurrency for Cilk.  United States: N. p., 2015. 
        Web.  doi:10.1145/2807591.2807597.

Copy to clipboard


                    Aga, Shaizeen D., Krishnamoorthy, Sriram, & Narayanasamy, Satish. CilkSpec: Optimistic Concurrency for Cilk.  United States.  https://doi.org/10.1145/2807591.2807597

Copy to clipboard


                    Aga, Shaizeen D., Krishnamoorthy, Sriram, and Narayanasamy, Satish. 2015.  
        "CilkSpec: Optimistic Concurrency for Cilk".  United States.  https://doi.org/10.1145/2807591.2807597.

Copy to clipboard


                    
@article{osti_1236312,

  title        = {CilkSpec: Optimistic Concurrency for Cilk},

  author       = {Aga, Shaizeen D. and Krishnamoorthy, Sriram and Narayanasamy, Satish},

  abstractNote = {Recursive parallel programming models such as Cilk strive to simplify the task of parallel programming by enabling a simple divide-and-conquer model of programming. This model is effective in re- cursively partitioning work into smaller parts and combining their results. However, recursive work partitioning can impose additional constraints on concurrency than is implied by the true dependences in a program. In this paper, we present a speculation-based approach to alleviate the concurrency constraints imposed by such recursive parallel programs. We design a runtime infrastructure that supports speculative execution and a predictor to accurately learn and identify opportunities to relax extraneous concurrency constraints. Experimental evaluation demonstrates that speculative relaxation of concurrency constraints can deliver gains of up to 1.6× on 30 cores over baseline Cilk.},

  doi          = {10.1145/2807591.2807597},

  url          = {https://www.osti.gov/biblio/1236312},
  journal      = {},
number       = ,

  volume       = ,

  place        = {United States},

  year         = {Sun Nov 15 00:00:00 EST 2015},

  month        = {Sun Nov 15 00:00:00 EST 2015}

}

Copy to clipboard

Conference:

https://doi.org/10.1145/2807591.2807597

Other availability

Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share:

Export Metadata

Save to My Library

Similar records in OSTI.GOV collections:

High Performance Computing Based Parallel HIearchical Modal Association Clustering (HPAR HMAC)

Software Patlolla, Dilip; Surendran Nair, Sujithkumar; Graves, Daniel

For many applications, clustering is a crucial step in order to gain insight into the makeup of a dataset. The best approach to a given problem often depends on a variety of factors, such as the size of the dataset, time restrictions, and soft clustering requirements. The HMAC algorithm seeks to combine the strengths of 2 particular clustering approaches: model-based and linkage-based clustering. One particular weakness of HMAC is its computational complexity. HMAC is not practical for mega-scale data clustering. For high-definition imagery, a user would have to wait months or years for a result; for a 16-megapixel image, themore »« less
View Software
Cache Locality Optimization for Recursive Programs

Conference Lifflander, Jonathan; Krishnamoorthy, Sriram

We present an approach to optimize the cache locality for recursive programs by dynamically splicing--recursively interleaving--the execution of distinct function invocations. By utilizing data effect annotations, we identify concurrency and data reuse opportunities across function invocations and interleave them to reduce reuse distance. We present algorithms that efficiently track effects in recursive programs, detect interference and dependencies, and interleave execution of function invocations using user-level (non-kernel) lightweight threads. To enable multi-core execution, a program is parallelized using a nested fork/join programming model. Our cache optimization strategy is designed to work in the context of a random work stealing scheduler. Wemore »« less
https://doi.org/10.1145/3140587.3062385
WE-A-16A-01: International Medical Physics Symposium: Increasing Access to Medical Physics Education/Training and Research Excellence

Journal Article Bortfeld, T; Ngoma, T; Odedina, F; ... - Medical Physics

In response to a world in which cancer is a growing global health challenge, there is now a greater need for US Medical Physicists and other Radiation Oncology professionals across institutions to work together and be more globally engaged in the fight against cancer. There are currently many opportunities for Medical Physicists to contribute to alleviating this pressing need, especially in helping enhance access to Medical Physics Education/training and Research Excellence across international boundaries, particularly for low and middle-income countries (LMIC), which suffer from a drastic shortage of accessible knowledge and quality training programs in radiotherapy. Many Medical Physicists aremore »« less
https://doi.org/10.1118/1.4889368
DoE Phase 1 Final Technical Report for MokaBlox

Technical Report Sadoghi, Mohammad; Zhao, Dongfang; McGregor, Kirk

MokaBlox is a university spinoff company building the next generation energy-conscious blockchain technology specifically designed for HPC environments to ensure a decentralized and democratic model for a provable and accountable cyber security and privacy. Blockchain infrastructure can be and is starting to be a provable and accountable model of cybersecurity that is valuable especially in terms of chronically under-appreciated resiliency, performance, and energy efficiency. This project aimed to develop an energy-conscious blockchain technology targeted for HPC cybersecurity. We have reimagined HPC cybersecurity and privacy model as built upon a decentralized and democractic computation model of blockchain that is not onlymore »« less
Data Locality Enhancement of Dynamic Simulations for Exascale Computing (Final Report)

Technical Report Shen, Xipeng

The development of modern processors exhibits two trends that complicate the optimizations of modern software. The first is the increasing sensitivity of processors' throughput to irregularities in computation. With more processors produced through a massive integration of simple cores, future systems will increasingly favor regular data-level parallel computations, but deviate from the needs of applications with complex patterns. Some evidences are already shown on Graphic Processing Units (GPU): Irregular data accesses (e.g., indirect references A[D[i]]) and conditional branches are limiting many GPU applications' performance at a level an order of magnitude lower than the peak of GPU. The second hardwaremore »« less
https://doi.org/10.2172/1576175

Full Text Available

Similar Records