FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3201951111> ?p ?o ?g. }

• W3201951111 abstract "Emerging edge computing platforms often contain machine learning (ML) accelerators that can accelerate inference for a wide range of neural network (NN) models. These models are designed to fit within the limited area and energy constraints of the edge computing platforms, each targeting various applications (e.g., face detection, speech recognition, translation, image captioning, video analytics). To understand how edge ML accelerators perform, we characterize the performance of a commercial Google Edge TPU, using 24 Google edge NN models (which span a wide range of NN model types) and analyzing each NN layer within each model. We find that the Edge TPU suffers from three major shortcomings: (1) it operates significantly below peak computational throughput, (2) it operates significantly below its theoretical energy efficiency, and (3) its memory system is a large energy and performance bottleneck. Our characterization reveals that the one-size-fits-all, monolithic design of the Edge TPU ignores the high degree of heterogeneity both across different NN models and across different NN layers within the same NN model, leading to the shortcomings we observe. We propose a new acceleration framework called Mensa. Mensa incorporates multiple heterogeneous edge ML accelerators (including both on-chip and near-data accelerators), each of which caters to the characteristics of a particular subset of NN models and layers. During NN inference, for each NN layer, Mensa decides which accelerator to schedule the layer on, taking into account both the optimality of each accelerator for the layer and layer-to-layer communication costs. Averaged across all 24 Google edge NN models, Mensa improves energy efficiency and throughput by 3.0x and 3.1x over the Edge TPU, and by 2.4x and 4.3x over Eyeriss~v2, a state-of-the-art accelerator." @default.
• W3201951111 created "2021-10-11" @default.
• W3201951111 creator A5003609686 @default.
• W3201951111 creator A5016220809 @default.
• W3201951111 creator A5029512221 @default.
• W3201951111 creator A5036666743 @default.
• W3201951111 creator A5050695684 @default.
• W3201951111 creator A5054097137 @default.
• W3201951111 creator A5065175808 @default.
• W3201951111 creator A5091016237 @default.
• W3201951111 date "2021-09-29" @default.
• W3201951111 modified "2023-09-23" @default.
• W3201951111 title "Google Neural Network Models for Edge Devices: Analyzing and Mitigating Machine Learning Inference Bottlenecks" @default.
• W3201951111 cites W1485009520 @default.
• W3201951111 cites W1498436455 @default.
• W3201951111 cites W1499864241 @default.
• W3201951111 cites W1508856789 @default.
• W3201951111 cites W1514535095 @default.
• W3201951111 cites W1546771929 @default.
• W3201951111 cites W1568514080 @default.
• W3201951111 cites W1689711448 @default.
• W3201951111 cites W1810943226 @default.
• W3201951111 cites W1828163288 @default.
• W3201951111 cites W1849277567 @default.
• W3201951111 cites W1895577753 @default.
• W3201951111 cites W1905882502 @default.
• W3201951111 cites W1924770834 @default.
• W3201951111 cites W1934184906 @default.
• W3201951111 cites W1934410531 @default.
• W3201951111 cites W1947481528 @default.
• W3201951111 cites W1989831790 @default.
• W3201951111 cites W2005708641 @default.
• W3201951111 cites W2064675550 @default.
• W3201951111 cites W2067523571 @default.
• W3201951111 cites W2097117768 @default.
• W3201951111 cites W2101926813 @default.
• W3201951111 cites W2112796928 @default.
• W3201951111 cites W2116261113 @default.
• W3201951111 cites W2116435618 @default.
• W3201951111 cites W2117539524 @default.
• W3201951111 cites W2130942839 @default.
• W3201951111 cites W2131218797 @default.
• W3201951111 cites W2152839228 @default.
• W3201951111 cites W2154579312 @default.
• W3201951111 cites W2160428323 @default.
• W3201951111 cites W2172140247 @default.
• W3201951111 cites W2194775991 @default.
• W3201951111 cites W2279098554 @default.
• W3201951111 cites W2289252105 @default.
• W3201951111 cites W2373570000 @default.
• W3201951111 cites W2414912620 @default.
• W3201951111 cites W2466675884 @default.
• W3201951111 cites W2473640056 @default.
• W3201951111 cites W2516141709 @default.
• W3201951111 cites W2525778437 @default.
• W3201951111 cites W2545177271 @default.
• W3201951111 cites W2565305208 @default.
• W3201951111 cites W2573587735 @default.
• W3201951111 cites W2585720638 @default.
• W3201951111 cites W2591013610 @default.
• W3201951111 cites W2594492285 @default.
• W3201951111 cites W2605258629 @default.
• W3201951111 cites W2605347906 @default.
• W3201951111 cites W2606722458 @default.
• W3201951111 cites W2612076670 @default.
• W3201951111 cites W2612445135 @default.
• W3201951111 cites W2616915022 @default.
• W3201951111 cites W2618530766 @default.
• W3201951111 cites W2625457103 @default.
• W3201951111 cites W2626991402 @default.
• W3201951111 cites W2752813595 @default.
• W3201951111 cites W2762910930 @default.
• W3201951111 cites W2786562507 @default.
• W3201951111 cites W2789554134 @default.
• W3201951111 cites W2790925711 @default.
• W3201951111 cites W2796625795 @default.
• W3201951111 cites W2883780447 @default.
• W3201951111 cites W2883929540 @default.
• W3201951111 cites W2884687835 @default.
• W3201951111 cites W2888243885 @default.
• W3201951111 cites W2919115771 @default.
• W3201951111 cites W2921796727 @default.
• W3201951111 cites W2931743911 @default.
• W3201951111 cites W2935331687 @default.
• W3201951111 cites W2944681516 @default.
• W3201951111 cites W2945146780 @default.
• W3201951111 cites W2950635152 @default.
• W3201951111 cites W2952453038 @default.
• W3201951111 cites W2953212265 @default.
• W3201951111 cites W2962760690 @default.
• W3201951111 cites W2962835968 @default.
• W3201951111 cites W2962949934 @default.
• W3201951111 cites W2963125010 @default.
• W3201951111 cites W2963414781 @default.
• W3201951111 cites W2963451564 @default.
• W3201951111 cites W2963510045 @default.
• W3201951111 cites W2964095005 @default.
• W3201951111 cites W2976697034 @default.
• W3201951111 cites W2979577066 @default.
• W3201951111 cites W2980104813 @default.

• next