A Reality Check on Inference at Mobile Networks Edge

Cartas, Alejandro; Kocour, Martin; Leontiadis, Ilias; Luque, Jordi; Sastry, Nishanth; Nunez-Martinez, Jose; Perino, Diego; Segura, Carlos

doi:10.1145/3301418.3313946

Cited by 28 publications

(18 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recent investigations report performance of preliminary deployment of 5G in the realworld to be sub-optimal [49,71]. Likewise, Hadzic et al [31] and Cartas et al [12] find that latency gains for accessing edge server colocated with an LTE basestation is minimal compared to accessing a datacenter located ≈ 1000 km away. While the "true" gains of 5G are yet to be seen, considering supporting strict MTP thresholds, even with edge servers located at basestations, seems uncertain.…”

Section: Discussionmentioning

confidence: 99%

“…The more pervasive deployment edge needs, the lower the likely processing capabilities of individual edge servers become. It is thus quite possible that despite extensive edge deployment, faster processing and availability of specialized hardware (like GPUs) offered by the cloud may far exceed the network latency gains from deploying applications at the edge [12].…”

Section: Discussionmentioning

confidence: 99%

“…Second, modern smartphones come equipped with considerable processing power, including specialized chipsets, enabling them to process complex tasks, such as AR. On the other hand, last-mile access being the latency bottleneck over (cellular) wireless remains true [12,31]. While new technologies, such as 5G, show promise, initial tests report their performance to be deficient in practice [49,71].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Pruning Edge Research with Latency Shears

Mohan

Corneo

Zavodovski

et al. 2020

Proceedings of the 19th ACM Workshop on Hot Topics in Networks

View full text Add to dashboard Cite

Edge computing has gained attention from both academia and industry by pursuing two significant challenges: 1) moving latency critical services closer to the users, 2) saving network bandwidth by aggregating large flows before sending them to the cloud. While the rationale appeared sound at its inception almost a decade ago, several current trends are impacting it. Clouds have spread geographically reducing end-user latency, mobile phones' computing capabilities are improving, and network bandwidth at the core keeps increasing. In this paper, we scrutinize edge computing, examining its outlook and future in the context of these trends. We perform extensive client-to-cloud measurements using RIPE Atlas, and show that latency reduction as motivation for edge is not as persuasive as once believed; for most applications the cloud is already "close enough" for majority of the world's population. This implies that edge computing may only be applicable for certain application niches, as opposed to a general-purpose solution.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Pruning Edge Research with Latency Shears

Mohan

Corneo

Zavodovski

et al. 2020

Proceedings of the 19th ACM Workshop on Hot Topics in Networks

View full text Add to dashboard Cite

show abstract

“…Towards this direction, binary networks [8,22] seem to o er a promising alternative for maximal compression, but currently require specialized hardware support to exploit the speedup potential. Finally, it is possible to dynamically o oad computation from device to the edge or cloud, in order to facilitate computation and minimize latency [5].…”

Section: Discussionmentioning

confidence: 99%

EmBench

Almeida

Laskaridis

Leontiadis

et al. 2019

The 3rd International Workshop on Deep Learning for Mobile Systems and Applications

Self Cite

View full text Add to dashboard Cite

In recent years, advances in deep learning have resulted in unprecedented leaps in diverse tasks spanning from speech and object recognition to context awareness and health monitoring. As a result, an increasing number of AI-enabled applications are being developed targeting ubiquitous and mobile devices. While deep neural networks (DNNs) are getting bigger and more complex, they also impose a heavy computational and energy burden on the host devices, which has led to the integration of various specialized processors in commodity devices. Given the broad range of competing DNN architectures and the heterogeneity of the target hardware, there is an emerging need to understand the compatibility between DNN-platform pairs and the expected performance bene ts on each platform. This work attempts to demystify this landscape by systematically evaluating a collection of state-of-the-art DNNs on a wide variety of commodity devices. In this respect, we identify potential bottlenecks in each architecture and provide important guidelines that can assist the community in the co-design of more e cient DNNs and accelerators.

show abstract

“…While cost and power consumption are the main limiting factors for onboard computation, the long latency and the massive data transmission are the bottlenecks for Cloud based processing [26]. The round trip time from a mobile client to a cloud center may easily be longer than 100 ms [27]. Moreover, such latency highly depends on the wireless channel condition, the bandwidth of the network, and traffic congestion, so real-time processing and reliability cannot be ensured.…”

Section: Introductionmentioning

confidence: 99%

Mobile Edge Intelligence and Computing for the Internet of Vehicles

2020

View full text Add to dashboard Cite

The Internet of Vehicles (IoV) is an emerging paradigm, driven by recent advancements in vehicular communications and networking. Advances in research can now provide reliable communication links between vehicles, via vehicle-to-vehicle communications, and between vehicles and roadside infrastructures, via vehicle-to-infrastructure communications. Meanwhile, the capability and intelligence of vehicles are being rapidly enhanced, and this will have the potential of supporting a plethora of new exciting applications, which will integrate fully autonomous vehicles, the Internet of Things (IoT), and the environment. These trends will bring about an era of intelligent IoV, which will heavily depend upon communications, computing, and data analytics technologies. To store and process the massive amount of data generated by intelligent IoV, onboard processing and Cloud computing will not be sufficient, due to resource/power constraints and communication overhead/latency, respectively. By deploying storage and computing resources at the wireless network edge, e.g., radio access points, the edge information system (EIS), including edge caching, edge computing, and edge AI, will play a key role in the future intelligent IoV. Such system will provide not only low-latency content delivery and computation services, but also localized data acquisition, aggregation and processing. This article surveys the latest development in EIS for intelligent IoV. Key design issues, methodologies and hardware platforms are introduced. In particular, typical use cases for intelligent vehicles are illustrated, including edge-assisted perception, mapping, and localization. In addition, various open research problems are identified.

show abstract

A Reality Check on Inference at Mobile Networks Edge

Cited by 28 publications

References 20 publications

Pruning Edge Research with Latency Shears

Pruning Edge Research with Latency Shears

EmBench

Mobile Edge Intelligence and Computing for the Internet of Vehicles

Contact Info

Product

Resources

About