GPUSync: A Framework for Real-Time GPU Management

Elliott, Glenn A.; Ward, Bryan C.; Anderson, James H.

doi:10.1109/rtss.2013.12

Cited by 110 publications

(65 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although this software stack enables pipelined multithreaded execution, applications remain at the mercy of the stock GPU software stack with respect to GPU resource arbitration and scheduling. As we have demonstrated in prior work [5,6], this is insufficient for supporting predictable real-time execution of GPU-using applications.…”

Section: Adding Real-time Support To Visionworkmentioning

confidence: 97%

“…First, we report on our efforts to implement a variant of OpenVX that is amenable to real-time analysis. With support from NVIDIA, we adapted an alphaversion of an NVIDIA OpenVX implementation called VisionWorks ® [3] to run atop PGM RT (a graph-based middleware developed by our group previously [4]), GPUSync (a real-time GPU management framework developed by our group previously [5,6]), and LITMUS RT (a real-time Linux extension jointly maintained by our group and MPI [7]). 4 Second, we explain how to transform OpenVX graphs to eliminate cycles due to back edges and support pipelining.…”

Section: Introductionmentioning

confidence: 99%

“…In this section, we present a brief overview of GPUSync [5,6,8] and discuss OpenVX in greater detail. GPUSync is a GPU management framework designed to be used on multicore platforms where multiple GPUs may be present.…”

Section: Introductionmentioning

confidence: 99%

“…There are also higher-level data objects common to computer vision algorithms-these include histograms, image pyramids, and lookup tables. 5 The programmer constructs a computer vision algorithm by instantiating primitives as nodes and data objects as parameters. The programmer binds parameters to node inputs and outputs.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Supporting Real-Time Computer Vision Workloads Using OpenVX on Multicore+GPU Platforms

Elliott

Yang

Anderson

2015

2015 IEEE Real-Time Systems Symposium

Self Cite

View full text Add to dashboard Cite

Abstract-In the automotive industry, there is currently great interest in supporting driver-assist and autonomouscontrol features that utilize vision-based sensing through cameras. The usage of graphics processing units (GPUs) can potentially enable such features to be supported in a cost-effective way, within an acceptable size, weight, and power envelope. OpenVX is an emerging standard for supporting computer vision workloads. OpenVX uses a graph-based software architecture designed to enable efficient computation on heterogeneous platforms, including those that use accelerators like GPUs. Unfortunately, in settings where real-time constraints exist, the usage of OpenVX poses certain challenges. For example, pipelining is difficult to support and processing graphs may have cycles. In this paper, graph transformation techniques are presented that enable these issues to be circumvented. Additionally, a case-study evaluation is presented involving an OpenVX implementation in which these techniques are applied. This OpenVX implementation runs atop a previously developed GPU-management framework called GPUSync. In this case study, the usage of GPUSync's GPU management techniques along with the proposed graph transformations enabled computer vision workloads specified using OpenVX to be supported in a predictable way.

show abstract

Section: Adding Real-time Support To Visionworkmentioning

confidence: 97%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Supporting Real-Time Computer Vision Workloads Using OpenVX on Multicore+GPU Platforms

Elliott

Yang

Anderson

2015

2015 IEEE Real-Time Systems Symposium

Self Cite

View full text Add to dashboard Cite

show abstract

“…In [EA12], the same authors present two analysis methods that permit the integration of GPUs in soft real-time multiprocessor systems. They follow up their work in [EWA13], where they propose and analyze GPUSync, a highly configurable real-time GPU management framework.…”

Section: Real-time Computing On Gpusmentioning

confidence: 99%

Hardware/Software Codesign of Embedded Systems with Reconfigurable and Heterogeneous Platforms

Lifa¹

2015

View full text Add to dashboard Cite

M odern applications running on today's embedded systems have very high requirements. Most often, these requirements have many dimensions: the applications need high performance as well as flexibility, energyefficiency as well as real-time properties, fault tolerance as well as low cost. In order to meet these demands, the industry is adopting architectures that are more and more heterogeneous and that have reconfiguration capabilities. Unfortunately, this adds to the complexity of designing streamlined applications that can leverage the advantages of such architectures.In this context, it is very important to have appropriate tools and design methodologies for the optimization of such systems. This thesis addresses the topic of hardware/software codesign and optimization of adaptive real-time systems implemented on reconfigurable and heterogeneous platforms. We focus on performance enhancement for dynamically reconfigurable FPGA-based systems, energy minimization in multi-mode real-time systems implemented on heterogeneous platforms, and codesign techniques for fault-tolerant systems.The solutions proposed in this thesis have been validated by extensive experiments, ranging from computer simulations to proof of concept implementations on real-life platforms. The results have confirmed the importance of the addressed aspects and the applicability of our techniques for design optimization of modern embedded systems. v vi Populärvetenskaplig Sammanfattning I dag är inbyggda system vanligt förekommande och deras antal fortsätter att öka. De används inom en mängd olika domäner, t.ex. hemelektronik, fordonsindustri, flygelektronik, medicin, etc., och de hittas nästan överallt omkring oss: från våra telefoner, bärbara datorer och tvättmaskiner till våra bilar. De applikationer som körs på dessa system har flera olika ökande krav: högprestanda, energieffektivitet, flexibilitet, feltolerans, realtidsegenskaper, och naturligtvis låg kostnad. För att kunna uppfylla dessa krav har industrin anammat arkitekturer som är mer och mer heterogena och som har omkonfigureringsmöjligheter. Olyckligtvis ökar detta komplexiteten när man ska utforma effektiva inbyggda system. I detta sammanhang blir det av yttersta betydelse att utveckla effektiva optimeringsmetoder och verktyg.Dagens applikationer består av en blandning av mjukvarukomponenter som har mycket olika energi-och prestandaegenskaper beroende på vilka hårdvaruenheter där de körs på, vilket gör dem lämpliga för heterogena plattformar. En heterogen plattform består av olika typer av hårdvaruen-heter, var och en med sina egenskaper, som riktar sig till vissa tillämpn-ingsområden. Under det senaste decenniet har utvecklingen av rekonfigurerbara hårdvarutekniker accelererat, vilket bidrar till den ökande populariteten för fältprogrammerbara grindmatriser (FPGA). Idag, ger hård-varutillverkare stöd för partiell dynamisk omkonfigurering; detta innebär att delar av en FPGA kan konfigureras vid run-time, medan andra delar förblir fullt fungerande. Dessa tekniska framsteg har...

show abstract

Assessing and Improving the Suitability of Model-Based Design for GPU-Accelerated Railway Control Systems

Calderon

Kosmidis

Nicolás

et al. 2021

Architecture of Computing Systems

View full text Add to dashboard Cite

Model-Based Design (MBD) is widely used for the design and simulation of electric traction control systems in the railway industry. Moreover, similar to other transportation industries, railway is moving towards the consolidation of multiple computing systems on fewer and more powerful ones, aiming for the reduction of Size, Weight and Power (SWaP). In that regard, Graphics Processing Units (GPUs) are increasingly considered by critical systems engineers, seeking to satisfy their ever increasing performance requirements. Recently, MBD tools have been enhanced with GPU code generation capabilities for machine learning acceleration, however, there is no indication whether these tools are ready for the design of time-sensitive systems. In this paper we analyse the suitability of commercial MBD toolsets by designing and deploying a model-based parallel control case study on embedded GPU platforms. While our results show promising feasibility evidence, they also reveal shortcomings which should be addressed before these toolsets become fit for developing critical systems. We propose certain improvements that have to be incorporated in these tools to achieve this goal. By implementing our proposals in the generated code, we experimentally show their effectiveness on two NVIDIA-based embedded GPUs.

show abstract

GPUSync: A Framework for Real-Time GPU Management

Cited by 110 publications

References 27 publications

Supporting Real-Time Computer Vision Workloads Using OpenVX on Multicore+GPU Platforms

Supporting Real-Time Computer Vision Workloads Using OpenVX on Multicore+GPU Platforms

Hardware/Software Codesign of Embedded Systems with Reconfigurable and Heterogeneous Platforms

Assessing and Improving the Suitability of Model-Based Design for GPU-Accelerated Railway Control Systems

Contact Info

Product

Resources

About