Abstract Machine Models and Proxy Architectures for Exascale Computing

Ang, James A.; Barrett, Richard Frederick; Burke, Daniel J.; Chan, Cy; Cook, Jeanine; Daley, Christopher; Donofrio, David; Hammond, Simon David; Hemmert, Karl Scott; Hoekstra, Robert J.; Ibrahim, Khaled Z.; Kelly, Suzanne M.; Le, Hoang M.; Leung, Vitus J.; Resnick, David Richard; Rodrigues, Arun F.; Shalf, John; Stark, Dylan; Unat, D.; Wright, Nicholas; Renae, Voskuilen, Gwendolyn

doi:10.2172/1561498

Cited by 15 publications

(27 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We extend our framework to consider the case where each platform node is equipped with a (private) burst buffer. Burst buffer integration is being considered in many future HPC architectures for scalable distributed storage mechanisms and to reduce I/O contention [12,13]. Here, we study burst buffers as a mechanism to mitigate CR I/O contention from concurrent application instances.…”

Section: Burst Buffersmentioning

confidence: 99%

Checkpointing Strategies for Shared High-Performance Computing Platforms

Hérault

Robert

Bouteiller

et al. 2019

IJNC

View full text Add to dashboard Cite

Input/output (I/O) from various sources often contend for scarcely available bandwidth. For example, checkpoint/restart (CR) protocols can help to ensure application progress in failureprone environments. However, CR I/O alongside an application's normal, requisite I/O can increase I/O contention and might negatively impact performance. In this work, we consider different aspects (system-level scheduling policies and hardware) that optimize the overall performance of concurrently executing CR-based applications that share I/O resources. We provide a theoretical model and derive a set of necessary constraints to minimize the global waste on a given platform. Our results demonstrate that Young/Daly's optimal checkpoint interval, despite providing a sensible metric for a single, undisturbed application, is not sufficient to optimally address resource contention at scale. We show that by combining optimal checkpointing periods with contention-aware system-level I/O scheduling strategies, we can significantly improve overall application performance and maximize the platform throughput. Finally, we evaluate how specialized hardware, namely burst buffers, may help to mitigate the I/O contention problem. Overall, these results provide critical analysis and direct guidance on how to design efficient, CR ready, large-scale platforms without a large investment in the I/O subsystem.

show abstract

Section: Burst Buffersmentioning

confidence: 99%

Checkpointing Strategies for Shared High-Performance Computing Platforms

Hérault

Robert

Bouteiller

et al. 2019

IJNC

View full text Add to dashboard Cite

show abstract

“…A key feature of DIY2 is its ability to move data seamlessly across the memory/storage hierarchy. This feature is becoming increasingly important on HPC architectures with limited memory capacity per core but increasing number of levels of memory/storage; for example, the U.S. Department of Energy (DOE) leadership computing architectures [2].…”

Section: Out-of-core Movementmentioning

confidence: 99%

“…The Center for Exascale Simulation of Advanced Reactors (CESAR) [40], one of three co-design centers funded by the DOE, contains a suite of data analysis benchmarks called cian. 2 It exercises the main communication patterns used in many distributed-memory data analysis algorithms: neighbor exchange, merge reduction, swap reduction, and parallel sorting. The message sizes and numbers of blocks are configurable in the cian mini-app; hence, one may test a particular regime such as latency-or bandwidth-bound communication at a variety of system sizes.…”

Section: Benchmark Applicationsmentioning

confidence: 99%

“…For example, burst buffers already exist in small prototype instances, and next generation of supercomputers will include them in production. Additional levels of fast/slow memory [2] and near/far disks create opportunities while complicating algorithm design. A related issue is that the number of cores is rapidly increasing; hundreds per node is now the norm.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Block-Parallel Data Analysis with DIY2

Morozov

Peterka

2017

View full text Add to dashboard Cite

DIY2 is a programming model and runtime for block-parallel analytics on distributedmemory machines. Its main abstraction is block-structured data parallelism: data are decomposed into blocks; blocks are assigned to processing elements (processes or threads); computation is described as iterations over these blocks, and communication between blocks is defined by reusable patterns. By expressing computation in this general form, the DIY2 runtime is free to optimize the movement of blocks between slow and fast memories (disk and flash vs. DRAM) and to concurrently execute blocks residing in memory with multiple threads. This enables the same program to execute in-core, out-of-core, serial, parallel, single-threaded, multithreaded, or combinations thereof. This paper describes the implementation of the main features of the DIY2 programming model and optimizations to improve performance. DIY2 is evaluated on benchmark test cases to establish baseline performance for several common patterns and on larger complete analysis codes running on large-scale HPC machines.

show abstract

“…Motivating applications and their requirements, 2. Data structure and layout abstractions, 3. Language and compiler support for data locality, 4.…”

Section: Summary Of Findings and Recommendationsmentioning

confidence: 99%

Programming Abstractions for Data Locality

Tate¹,

Kamil

Dubey

et al. 2014

Self Cite

View full text Add to dashboard Cite

The goal of the workshop and this report is to identify common themes and standardize concepts for locality-preserving abstractions for exascale programming models. Current software tools are built on the premise that computing is the most expensive component, we are rapidly moving to an era that computing is cheap and massively parallel while data movement dominates energy and performance costs. In order to respond to exascale systems (the next generation of high performance computing systems), the scientific computing community needs to refactor their applications to align with the emerging data-centric paradigm. Our applications must be evolved to express information about data locality. Unfortunately current programming environments offer few ways to do so. They ignore the incurred cost of communication and simply rely on the hardware cache coherency to virtualize data movement. With the increasing importance of task-level parallelism on future systems, task models have to support constructs that express data locality and affinity. At the system level, communication libraries implicitly assume all the processing elements are equidistant to each other. In order to take advantage of emerging technologies, application developers need a set of programming abstractions to describe data locality for the new computing ecosystem. The new programming paradigm should be more data centric and allow to describe how to decompose and how to layout data in the memory. Fortunately, there are many emerging concepts such as constructs for tiling, data layout, array views, task and thread affinity, and topology aware communication libraries for managing data locality. There is an opportunity to identify commonalities in strategy to enable us to combine the best of these concepts to develop a comprehensive approach to expressing and managing data locality on exascale programming systems. These programming model abstractions can expose crucial information about data locality to the compiler and runtime system to enable performance-portable code. The research question is to identify the right level of abstraction, which includes techniques that range from template libraries all the way to completely new languages to achieve this goal.

show abstract

Abstract Machine Models and Proxy Architectures for Exascale Computing

Cited by 15 publications

References 8 publications

Checkpointing Strategies for Shared High-Performance Computing Platforms

Checkpointing Strategies for Shared High-Performance Computing Platforms

Block-Parallel Data Analysis with DIY2

Programming Abstractions for Data Locality

Contact Info

Product

Resources

About