An evaluation of global address space languages

Coarfa, Cristian; Dotsenko, Yuri; Mellor-Crummey, John; Cantonnet, François; El‐Ghazawi, Tarek; Mohanti, Ashrujit; Yao, Yiyi; Chavarŕıa-Miranda, Daniel

doi:10.1145/1065944.1065950

Cited by 76 publications

(15 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For example, one can explicitly mark loops that have to be parallelized with annotations such as OpenMP [2] or OpenACC [3], or with an API such as OpenCL and Cuda. There are also approaches that rely on more fundamental language modifications such as Charm++ [4] or Partitioned Global Address Space (PGAS) languages such as UPC or Co-Array Fortran [5]. They embeds parallel constructs that aim to abstract actual machines.…”

Section: Target Specific Compilationmentioning

confidence: 99%

A low level component model easing performance portability of HPC applications

et al. 2013

View full text Add to dashboard Cite

International audienceScientific applications are getting increasingly complex, e.g., to improve their accuracy by taking into account more phenomena. Meanwhile, computing infrastructures are continuing their fast evolution. Thus, software engineering is becoming a major issue to offer ease of development, portability and maintainability while achieving high performance. Component based software engineering offers a promising approach that enables the manipulation of the software architecture of applications. However, existing models do not provide an adequate support for performance portability of HPC applications. This paper proposes a low level component model (L²C) that supports inter-component interactions for typical scenarios of high performance computing, such as process-local shared memory and function invocation (C++ and Fortran), MPI, and Corba. To study the benefits of using L²C, this paper walks through an example of stencil computation, i.e. a structured mesh Jacobi implementation of the 2D heat equation parallelized through domain decomposition. The experimental results obtained on the Grid'5000 testbed and on the Curie supercomputer show that L²C can achieve performance similar to that of native implementations, while easing performance portability

show abstract

Section: Target Specific Compilationmentioning

confidence: 99%

A low level component model easing performance portability of HPC applications

et al. 2013

View full text Add to dashboard Cite

show abstract

“…The idea of using qualifiers to distinguish between shared and private memory originated in SIMD array languages [13], and is used in PGAS languages such as Titanium [14], Co-array Fortran and Unified Parallel C [15]. Similar storage qualifiers are used by CUDA and OpenCL to specify data locations in accelerators with hierarchical memory.…”

Section: Related Workmentioning

confidence: 99%

Offload – Automating Code Migration to Heterogeneous Multicore Systems

Cooper

Dolinsky

Donaldson

et al. 2010

High Performance Embedded Architectures and Compilers

View full text Add to dashboard Cite

Abstract. We present Offload, a programming model for offloading parts of a C++ application to run on accelerator cores in a heterogeneous multicore system. Code to be offloaded is enclosed in an offload scope; all functions called indirectly from an offload scope are compiled for the accelerator cores. Data defined inside/outside an offload scope resides in accelerator/host memory respectively, and code to move data between memory spaces is generated automatically by the compiler. This is achieved by distinguishing between host and accelerator pointers at the type level, and compiling multiple versions of functions based on pointer parameter configurations using automatic call-graph duplication. We discuss solutions to several challenging issues related to call-graph duplication, and present an implementation of Offload for the Cell BE processor, evaluated using a number of benchmarks.

show abstract

“…Distributed shared memory (DSM) systems model global shared memory using distributed local memory. In order to facilitate better mapping to distributed memory, some DSM systems like UPC [Coarfa05] differentiate between local data and shared global data. Although both types of memory may be accessed through the uniform interface of pointer dereferencing, global data typically has a longer access time.…”

Section: Related Workmentioning

confidence: 99%

The Design of a Polymorphous Cognitive Agent Architecture (PCAA)

Amduka¹,

Russo²,

Jha³

et al. 2008

View full text Add to dashboard Cite

Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden to Washington Headquarters Service, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302, and PCAA is a dynamic, adaptive cognitive architecture that makes previously intractable approximation tasks tractable for NP-hard cognitive problems. PCAA consists of: linear composable cognitive agents, a cognitive mark-up language for cognitive behavior definition, a cognitive layer for derivation of cognitive services and specialized cognitive agents, and a next generation polymorphic hardware and software layer for runtime composition and instantiation of cognitive agents. PCAA is a dynamic, adaptive cognitive architecture that makes previously intractable approximation tasks tractable for NP-hard cognitive problems. PCAA consists of: linear composable cognitive agents, a cognitive mark-up language for cognitive behavior definition, a cognitive layer for derivation of cognitive services and specialized cognitive agents, and a next generation polymorphic hardware and software layer for runtime composition and instantiation of cognitive agents. SPONSOR/MONITOR'S ACRONYM(S) 9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)AFRL SUBJECT TERMSOur approach included a comprehensive concept study in the context of representative DoD challenge problems that have a clear and well-defined need for ACIP technology. PCAA application experiments demonstrated clear performance improvements over traditional computing architectures for cognitive processing for these applications. Our innovations include:• Dynamically composable hardware and software with linear scalability for cognitive processing across a massively parallel hardware fabric for real time autonomous systems. • A dynamically composed agent architecture that partitions reactive and predefined behaviors into linear lower level cognitive agents that tailor and adapt the overall behavior of the computing architecture to immediate mission needs.• Run-time derived cognitive virtual machines to partition cognitive processing to a new generation of computing run-time configured hardware and software to allow for dynamic cognitive computing reconfiguration required to achieve reactive processing.Our research was driven by DoD applications that have demonstrated needs for diverse cognitive processing that cannot be addressed by current computing hardware and software architectures. We demonstrated our end-to-end approach for two applications with direct DoD relevance: control of autonomous Unmanned Aerial Vehicles and Intelligence Analysis.ii

show abstract

An evaluation of global address space languages

Cited by 76 publications

References 13 publications

A low level component model easing performance portability of HPC applications

A low level component model easing performance portability of HPC applications

Offload – Automating Code Migration to Heterogeneous Multicore Systems

The Design of a Polymorphous Cognitive Agent Architecture (PCAA)

Contact Info

Product

Resources

About