9 Hits in 4.9 sec

On-Device Deep Learning Inference for System-on-Chip (SoC) Architectures

Tom Springer, Elia Eiroa-Lledo, Elizabeth Stevens, Erik Linstead
2021 Electronics  
The practicality of our scheduling framework was demonstrated by integrating it into a commercial real-time operating system (VxWorks) then running a typical deep learning image processing application  ...  The realization of machine learning, and deep learning, is being driven by the availability of specialized hardware, such as system-on-chip solutions, which provide some alleviation of constraints.  ...  on-chip learning.  ... 
doi:10.3390/electronics10060689 fatcat:efmsvkcg3rg3lk6tkk7txtx3uu

2020 Index IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems Vol. 39

2020 IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems  
., +, TCAD Nov. 2020 3968-3980 Scope-Aware Useful Cache Block Calculation for Cache-Related Pre-Emp- tion Delay Analysis With Set-Associative Data Caches.  ...  ., TCAD Aug. 2020 1739-1743 Capacity Augmentation Function for Real-Time Parallel Tasks With Constrained Deadlines Under GEDF Scheduling.  ...  Entropy-Directed Scheduling for FPGA High-Level Synthesis. Shen, M., +, TCAD Oct. 2020 2588 -2601 FLASH: Fast, Parallel, and Accurate Simulator for HLS.  ... 
doi:10.1109/tcad.2021.3054536 fatcat:wsw3olpxzbeclenhex3f73qlw4

Research on Three Dimensional Scheduling Model for Embedded Multi-Core System

Shujuan Huang, Yi'an Zhu, Bailin Liu, Feng Xiao
2018 Xibei Gongye Daxue Xuebao/Journal of Northwestern Polytechnical University  
当前多核实时任务的调 度算法主要分为全局调度算法( global scheduling al• gorithm) [4•6] 和分区调度算法 ( partitioned scheduling algorithm) [7•8] 2 大类。  ...  GEDF( global earliest deadline first ) [9] 和 PEDF ( partitioned earliest deadline first) [10] ,是 2 类调度算法的典型代表。 对于全局调 度算法来说,如果一个任务不允许抢占,而另外一个 任务必须执行,那么就要将该任务迁移到其他核上 执行 [11] 。 这种迁移开销比较大。  ...  Integrating CacheRelated Preemption Delay into GEDF Analysis for Multiprocessor Scheduling with OnChip Cache[ C] ∥Proceedings of the The 14th IEEE International Conference on Embedded Software and Systems  ... 
doi:10.1051/jnwpu/20183651020 fatcat:qabfmri575eejluqabmyrj3boy

A review of priority assignment in real-time systems

Robert I. Davis, Liliana Cucu-Grosjean, Marko Bertogna, Alan Burns
2016 Journal of systems architecture  
It is over 40 years since the first seminal work on priority assignment for real-time systems using fixed priority scheduling.  ...  Since then, huge progress has been made in the field of real-time scheduling with more complex models and schedulability analysis techniques developed to better represent and analyse real systems.  ...  The integration and analysis of overheads due to Cache Related Pre-emption Delays (CRPD) into fixed priority preemptive scheduling [4] also leads to an interesting and difficult to solve problem of priority  ... 
doi:10.1016/j.sysarc.2016.04.002 fatcat:ziodjnqp7jby5pskfspogugd4u

The Proceedings of First Work-in-Progress Session of The CSI International Symposium on Real-Time and Embedded Systems and Technologies [article]

Ali Behnoudfar, Seyyed Hossein Hosseini Zahani, Mojtaba Hatami, Mahmoud Naghibzadeh, Boshra Taheri, Fathieh Faghih, Mehdi Kargahi
2019 arXiv   pre-print
Each submission has been reviewed by 3 to 5 independent referees, for its quality, originality, contribution, clarity of presentation, and relevance to the symposium topics.  ...  This event has been organized by the School of Electrical and Computer Engineering at the University of Tehran, in conjunction with the Department of Computer Engineering at Sharif University of Technology  ...  ChronOS Real-Time Operating System ChronOS Linux is a best-effort real-time Linux kernel for chip multiprocessors (CMPs).  ... 
arXiv:1904.06995v1 fatcat:7lvriwiw5fd55njqithwclrxru

Real-Time Virtualization and Cloud Computing

Sisu Xi
The impact of cache has received significant attention in the context of one level scheduling [30, 59, 76] ; we defer integrating these insights into a two-level hierarchal scheduling to future work.  ...  We are actively working with Xen developers to integrate RT-Xen into the Xen mainstream to replace the legacy SEDF scheduler.  ... 
doi:10.7936/k7vm49bx fatcat:5gzhsh7zhjavvfndnj5yieoglq

Analysis and simulation of scheduling techniques for real-time embedded multi-core architectures [article]

Sanjib Das, Universität Stuttgart, Universität Stuttgart
As a consequence, the need for efficient yet predictable multi-core scheduling techniques is on the rise.  ...  The emphasis is on the degree of their usability on hard real-time systems, focusing on the scheduling techniques offering better processor affinity and the lower number of context switching.  ...  because of data reloading into cache.  ... 
doi:10.18419/opus-3460 fatcat:cry523mypnhcbjryucfxuc7v2i

WCET analysis and optimization for multi-core real-time systems

Timon Kelter, Technische Universität Dortmund, Technische Universität Dortmund
This non-local effect is called the Cache-Related Preemption Delay (CRPD).  ...  A more extreme example is the IBM SyNAPSE chip, according to IBM research the biggest chip that was ever built by IBM, which integrates 4096 cores on a single chip [Joh14] .  ...  B avg min ) and upper (B avg max ) loop bound and the number of context blocks V C τ for a context graph with unlimited call string length but without virtual unrolling.  ... 
doi:10.17877/de290r-7209 fatcat:yhl2jvlcvzh3rhc26azn4dnbre

Parallel Real-Time Scheduling for Latency-Critical Applications

Jing Li
All of these would benefit me for my academic career and the rest of life.  ...  They always believe in me, help me to get through difficulties and encourage me to reach for the stars. I have learned from them more than just research.  ...  Related Work on Hard Real-Time Systems Real-time scheduling for sequential tasks: Most prior work on real-time scheduling atop multiprocessors has concentrated on sequential tasks [54] .  ... 
doi:10.7936/k7b27tpk fatcat:hrhhktu54zdczjd73sfw6demdu