High-performance code optimizations for mobile devices

Afonso, Sergio; Acosta, Alejandro; Almeida, Francisco

doi:10.1007/s11227-018-2638-5

Cited by 3 publications

(3 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2019 A year later the same group has published a framework "TransferCL" [11] based on OpenCL that allows to use deep learning on mobile devices using the GPU. Afonso et al [12] have implemented a tiling optimization on GPUs of Android devices for tuning the execution of different algorithms. The same year Fasogbon et al [13] have implemented a Depth-Map algorithm on a mobile device using the GPU.…”

Section: A Contributionsmentioning

confidence: 99%

GPU backed Data Mining on Android Devices

Robert¹,

Plant²

2021

Preprint

View full text Add to dashboard Cite

Choosing an appropriate programming paradigm for high-performance computing on low-power devices can be useful to speed up calculations. Many Android devices have an integrated GPU and -although not officially supportedthe OpenCL framework can be used on Android devices for addressing these GPUs. OpenCL supports thread and data parallelism. Applications that use the GPU must account for the fact that they can be suspended by the user or the Android operating system at any moment.We have created a wrapper library that allows to use OpenCL on Android devices. Already written OpenCL programs can be executed with almost no modification. We have used this library to compare the performance of the DBSCAN and Kmeans algorithms on an integrated GPU of an Arm-v7 tablet with other single and multithreaded implementations on the same device. We have investigated which programming paradigm and language allows the best tradeoff between execution speed and energy consumption.Using the GPU for HPC on Android devices can help to carry out computationally intensive machine learning or data mining tasks in remote areas, under harsh environmental conditions and in areas where energy supply is an issue.

show abstract

Section: A Contributionsmentioning

confidence: 99%

GPU backed Data Mining on Android Devices

Robert¹,

Plant²

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…These are all programmed in various ways through software interfaces such as OpenCL, Renderscript and vendor-specific libraries, although there are frameworks that try to reduce the development complexity by unifying many of these interfaces [17], [18]. Writing high performance code for Android often relies on accelerators [19], and that is a case in which benchmarking is of paramount importance, so the interaction between these interfaces must be considered.…”

Section: Android Runtime Features a Common Runtime Featuresmentioning

confidence: 99%

Rancid: Reliable Benchmarking on Android Platforms

Afonso

Almeida

2020

IEEE Access

Self Cite

View full text Add to dashboard Cite

Benchmarking is an important step in the code optimization process that enables empirical performance evaluations in computer systems. Application profiling allows the detection of bottlenecks within the code, and benchmarking can be used to measure the effect of optimizations on performance and to compare implementations. However, obtaining reliable and reproducible performance metrics on modern mobile platforms is a complex task that is often overlooked. This is necessary to produce scientifically sound experiments. There are several factors that introduce noise on performance measurements. We identify and measure the most relevant set of these factors and design a methodology that enables more reliable performance benchmarking on Android platforms. We also describe our flexible benchmarking framework, Rancid, designed to transparently solve these problems. It enables application developers to quickly obtain reliable performance metrics for their code on a wide set of platforms. The evaluation of our methodology and framework shows an improvement on the behavior of results, successfully providing more precise and reproducible measurements on a range of devices and implementations.

show abstract

“…5 for further implementation details). Mobile SOCs are becoming more powerful and they are capable of executing compute-intensive applications, provided they are properly optimized [1,27].…”

Section: Introductionmentioning

confidence: 99%

Energy-efficient design of a presbyopia correction wearable powered by mobile GPUs and FPGAs

2022

View full text Add to dashboard Cite

This paper presents an energy-efficient design and evaluation of a novel portable device for the automatic correction of presbyopia in human eyes driven by the use of opto-electronic lenses and based on the dynamic pupil response of the subject. Due to the wearable nature of the proposed Dynamic Auto–Accommodation Glasses, in addition to the real-time requirement, an energy-efficient implementation is critical for the success of the device. In this work, the binocular pupil tracking of a subject, followed by the calculation of the eyes’ vergence, and the control of a pair of opto-electronic lenses are implemented and evaluated on several hardware platforms, including two mobile GPU/SoCs, a high-end FPGA, a low-cost FPGA, and a desktop GPU (as a reference point). The pupil tracking algorithm has been parallelized, applying different platform-specific optimizations for each case, to design a fast yet energy-efficient wearable. The hardware platforms have been evaluated to determine which one is the most appropriate for the presbyopia correction task. The experimental results show that the most energy-efficient platform is a mobile GPU (Samsung Exynos 8890) capable of processing frames at 0.016 Joules/frame, still allowing real-time processing (24 frames/sec).

show abstract

High-performance code optimizations for mobile devices

Cited by 3 publications

References 14 publications

GPU backed Data Mining on Android Devices

GPU backed Data Mining on Android Devices

Rancid: Reliable Benchmarking on Android Platforms

Energy-efficient design of a presbyopia correction wearable powered by mobile GPUs and FPGAs

Contact Info

Product

Resources

About