Using explicit platform descriptions to support programming of heterogeneous many-core systems

Sandrieser, Martin; Benkner, Siegfried; Pllana, Sabri

doi:10.1016/j.parco.2011.10.008

Cited by 26 publications

(13 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Additionally, the compute resource descriptor holds information about various PCI devices such as GPUs, Xeon Phi and other accelerators. Compute resource descriptors are specified using existing approaches -PDL [18] and hwloc [7]. In addition, we explicitly represent storage resource descriptors complementing compute resource descriptors with information about the The storage resources may comprise a set of integrated storage platforms annotated by their associated descriptors to map access endpoints for the execution engine.…”

Section: Resource Layermentioning

confidence: 99%

A containerized analytics framework for data and compute-intensive pipeline applications

Kaniovskyi

Koehler

Benkner

2017

Proceedings of the 4th ACM SIGMOD Workshop on Algorithms and Systems for MapReduce and Beyond

Self Cite

View full text Add to dashboard Cite

The joint effort of scientific collaborations and the expanding data market creates demand for high-performance and dataintensive analytics infrastructures that can exploit the potential of heterogeneous multi-core architectures with dynamic and scalable execution environments. Contemporary approaches focus on developing efficient parallel application models, but lack the flexibility of efficiently integrating and utilizing native or accelerator-based code. In this work, we illustrate a novel approach on mending this shortcoming and offering seamless application integration into a highly versatile execution infrastructure. The centerpiece is a framework of containerized execution units and management thereof for satisfying the diverse requirements of data analytics pipelines and its stages. Containers not only ease distribution and deployment of applications, but, more importantly enable an efficient synthesis of different stage implementation variants aimed towards exploiting heterogeneous computing resources. Consequently, this approach allows the infrastructure to utilize mainstream data and compute-intensive techniques and paradigms to achieve the goal of efficient pipeline execution. We present our approach in form of a requirement analysis, a multi-tier architecture description, and deployment scenarios based on our current prototype implementation.

show abstract

Section: Resource Layermentioning

confidence: 99%

A containerized analytics framework for data and compute-intensive pipeline applications

Kaniovskyi

Koehler

Benkner

2017

Proceedings of the 4th ACM SIGMOD Workshop on Algorithms and Systems for MapReduce and Beyond

Self Cite

View full text Add to dashboard Cite

show abstract

“…To deliver the envisaged performance analysis tools, we face the following challenges: (1) determining the computation of cost properties that are given by means of probabilistic distributions; (2) the inference of average cost in addition to the traditional worst-case cost; (3) take into account the underlying platform through a set of probabilistic parameters; (4) deal with massive and heterogeneous parallelism [14], [15], [16], [17] which is challenging for program analysis in general; and (5) develop multi-objective resource usage analyses and optimizations.…”

Section: B Static Analysis and Optimizationmentioning

confidence: 99%

Preparing HPC Applications for Exascale: Challenges and Recommendations

Ábrahám

Bekas

Brandić

et al. 2015

2015 18th International Conference on Network-Based Information Systems

Self Cite

View full text Add to dashboard Cite

Abstract-While the HPC community is working towards the development of the first Exaflop computer (expected around 2020), after reaching the Petaflop milestone in 2008 still only few HPC applications are able to fully exploit the capabilities of Petaflop systems. In this paper we argue that efforts for preparing HPC applications for Exascale should start before such systems become available. We identify challenges that need to be addressed and recommend solutions in key areas of interest, including formal modeling, static analysis and optimization, runtime analysis and optimization, and autonomic computing. Furthermore, we outline a conceptual framework for porting HPC applications to future Exascale computing systems and propose steps for its implementation.

show abstract

“…Furthermore, different flavors of heterogeneous systems are in use today, ranging from systems with general purpose GPUs to systems with FPGAs and heterogeneous CPUs (e.g., Cell). The PEP-PHER Platform Description Language (PDL) [189] allows modeling of key architectural features for these systems, both from hardware and software, and provides a unified interface for both programmers and tool-chains to access that information in a well-defined and explicit manner. We have implemented support for PEPPHER PDL in our framework to allow the component writer to specify constraints/hints by using key architectural properties which are resolved by our system either statically or dynamically.…”

Section: Support For Pdl (Platform Description Language)mentioning

confidence: 99%

Performance-aware component composition for GPU-based systems

Dastgeer¹

2014

View full text Add to dashboard Cite

This thesis adresses issues associated with efficiently programming modern heterogeneous GPU-based systems, containing multicore CPUs and one or more programmable Graphics Processing Units (GPUs). We use ideas from component-based programming to address programming, performance and portability issues of these heterogeneous systems. Specifically, we present three approaches that all use the idea of having multiple implementations for each computation; performance is achieved/retained either a) by selecting a suitable implementation for each computation on a given platform or b) by dividing the computation work across different implementations running on CPU and GPU devices in parallel.In the first approach, we work on a skeleton programming library (SkePU) that provides high-level abstraction while making intelligent implementation selection decisions underneath either before or during the actual program execution. In the second approach, we develop a composition tool that parses extra information (metadata) from XML files, makes certain decisions offline, and, in the end, generates code for making the final decisions at runtime. The third approach is a framework that uses source-code annotations and program analysis to generate code for the runtime library to make the selection decision at runtime. With a generic performance modeling API alongside program analysis capabilities, it supports online tuning as well as complex program transformations.These approaches differ in terms of genericity, intrusiveness, capabilities and knowledge about the program source-code; however, they all demonstrate usefulness of component programming techniques for programming GPU-based systems. With experimental evaluation, we demonstrate how all three approaches, although different in their own way, provide good performance on different GPU-based systems for a variety of applications.This work has been supported by two EU FP7 projects (PEP-PHER, EXCESS) and by SeRC. Populärvetenskaplig sammanfattningAtt få varje generation av datorer att fungera snabbareär viktigt för samhä-llets utveckling och tillväxt. Traditionellt hade de flesta datorer bara en general-purpose processor (den så kallade CPU:n) som bara kunde exekvera en beräkningsuppgift i taget. Under det senasteårtiondet har dock flerkärniga och mångkärniga processorer blivit vanliga, och datorer har också blivit mer heterogena. Ett modernt datorsystem innehåller vanligtvis flerän en CPU, tillsammans med specialprocessorer såsom grafikprocessorer (GPU:er) som ar anpassade för att kunna exekvera vissa typer av beräkningar effektivarë an CPU:er. Vi kallar ett sådant system med en eller flera GPU:er för ett GPU-baserat system. GPU:er i sådana system har sitt eget separata minne, och för att kunna köra en beräkning på en GPU så behöver man vanligtvis flytta all indata till GPU:ns minne och sedan hämta tillbaka resultatet när beräkningenär klar.Programmeringen av GPU-baserade systemär icke-trivialt av flera anledningar: (1) CPU:er och GPU:er kräver olika programmeringsexper...

show abstract

Using explicit platform descriptions to support programming of heterogeneous many-core systems

Cited by 26 publications

References 17 publications

A containerized analytics framework for data and compute-intensive pipeline applications

A containerized analytics framework for data and compute-intensive pipeline applications

Preparing HPC Applications for Exascale: Challenges and Recommendations

Performance-aware component composition for GPU-based systems

Contact Info

Product

Resources

About