Filters








961 Hits in 10.1 sec

Algorithms for Hierarchical and Semi-Partitioned Parallel Scheduling

Vincenzo Bonifaci, Gianlorenzo DAngelo, Alberto Marchetti-Spaccamela
2017 2017 IEEE International Parallel and Distributed Processing Symposium (IPDPS)  
For the makespan minimization objective, the model generalizes classical scheduling problems such as unrelated parallel machine scheduling, as well as novel ones such as semi-partitioned and clustered  ...  We propose a model for scheduling jobs in a parallel machine setting that takes into account the cost of migrations by assuming that the processing time of a job may depend on the specific set of machines  ...  indeed, in Example II.1, the original instance I of the semi-partitioned problem has an optimal makespan of 2, while the unrelated machine instance I u has an optimal makespan of 3.  ... 
doi:10.1109/ipdps.2017.22 dblp:conf/ipps/BonifaciDM17 fatcat:aqf5so6qvfalfi62ouxnpkj6g4

Multiprocessor Real-Time Scheduling with Hierarchical Processor Affinities

Vincenzo Bonifaci, Bjorn Brandenburg, Gianlorenzo DAngelo, Alberto Marchetti-Spaccamela
2016 2016 28th Euromicro Conference on Real-Time Systems (ECRTS)  
A notion of "strong arbitrary processor affinity scheduling" (strong APA scheduling) has been proposed; this notion avoids schedulability losses due to overly simple implementations of processor affinities  ...  Additionally, in the case of a bilevel affinity hierarchy and when job priorities are based on deadlines, we argue that the performance of our strong HPA scheduler, HPA-EDF, can be related to system optimality  ...  The authors also acknowledge Schloss Dagstuhl Seminar 16081 "Scheduling".  ... 
doi:10.1109/ecrts.2016.24 dblp:conf/ecrts/BonifaciBDM16 fatcat:cld2de5duja65f7r23eqnvigji

Multiprocessor real-time scheduling with arbitrary processor affinities: from practice to theory

Arpan Gujarati, Felipe Cerqueira, Björn B. Brandenburg
2014 Real-time systems  
It is shown that job-level fixed-priority scheduling with arbitrary processor affinities is strictly more general than global, clustered, and partitioned job-level fixed-priority scheduling combined.  ...  The Linux push and pull scheduler is studied as a reference implementation and two approaches for the schedulability analysis of hard real-time tasks with arbitrary processor affinity masks are presented  ...  To reduce pessimism in our experiments, in the response-time analysis for APA scheduling we included an optimization for uniprocessors affinity masks, which is explained next.  ... 
doi:10.1007/s11241-014-9205-9 fatcat:inmf6r42inektjhm63dxegnyti

Linux's Processor Affinity API, Refined: Shifting Real-Time Tasks Towards Higher Schedulability

Felipe Cerqueira, Arpan Gujarati, Bjorn B. Brandenburg
2014 2014 IEEE Real-Time Systems Symposium  
An empirical evaluation in terms of schedulability shows shifting to be effective, although performance naturally degrades if migration overheads are high.  ...  However, as the motivation to use processor affinities is highly application-specific, the two interests can conflict, i.e., the fixed, user-specified processor affinities chosen for non-schedulability  ...  Therefore, there is little incentive to use an APA scheduler at runtime if partitioning suffices to meet all design-time affinity restrictions.  ... 
doi:10.1109/rtss.2014.29 dblp:conf/rtss/CerqueiraGB14 fatcat:r324rc65nrdexa2lvzmh2v4f64

Variational Mixture of Normalizing Flows [article]

Guilherme G. P. Freitas Pires, Mário A. T. Figueiredo
2020 arXiv   pre-print
As will become clear, this model naturally lends itself to (multimodal) density estimation, semi-supervised learning, and clustering.  ...  The present work overcomes this by using normalizing flows as components in a mixture model and devising an end-to-end training procedure for such a model.  ...  This transformation is characterized by two arbitrary functions s(.) and t(.), as well as a mask that splits an input z of dimension D into two parts, z 1 and z 2 .  ... 
arXiv:2009.00585v1 fatcat:y7jkcwxi6ve55h2jzkdx4bvr6y

SOLO: Segmenting Objects by Locations [article]

Xinlong Wang, Tao Kong, Chunhua Shen, Yuning Jiang, Lei Li
2020 arXiv   pre-print
., semantic segmentation, it is the arbitrary number of instances that have made instance segmentation much more challenging.  ...  In order to predict a mask for each instance, mainstream approaches either follow the 'detect-thensegment' strategy as used by Mask R-CNN, or predict category masks first then use clustering techniques  ...  Acknowledgement We would like to thank Dongdong Yu and Enze Xie for the discussion about maskness and dice loss. We also thank Chong Xu and the ByteDance AI Lab team for technical support.  ... 
arXiv:1912.04488v3 fatcat:qjuuikzi3bcatamzsje7trcm3u

Multiprocessor Real-Time Locking Protocols: A Systematic Review [article]

Björn B. Brandenburg
2019 arXiv   pre-print
on maximum priority-inversion blocking, and matching asymptotically optimal locking protocols.  ...  A special focus is placed on the suspension-oblivious and suspension-aware analysis approaches for semaphore protocols, their respective notions of priority inversion, optimality criteria, lower bounds  ...  In fact, it can even be used in unchanged form under semi-partitioned scheduling or in the presence of tasks with arbitrary processor affinities (APAs) [95, 96] .  ... 
arXiv:1909.09600v1 fatcat:tmqcpiuxfbbd5jrcecgvoeanpm

A Schedulability Compatible Multiprocessor Resource Sharing Protocol -- MrsP

A. Burns, A.J. Wellings
2013 2013 25th Euromicro Conference on Real-Time Systems  
The form of the protocol is motivated by the desire to link the protocol with effective schedulability analysis.  ...  The protocol is general purpose, but is developed in this paper for partitioned fixed priority systems with the sporadic task model.  ...  SYSTEM AND TASK MODEL Our aim is to develop a general purpose protocol that is applicable to partitioned, semi-partitioned and globally scheduled systems using fixed priorities, EDF or any other designation  ... 
doi:10.1109/ecrts.2013.37 dblp:conf/ecrts/BurnsW13 fatcat:rxx7akpyezbzbpny5frumtz7mm

Discrete Energy Minimization, beyond Submodularity: Applications and Approximations [article]

Shai Bagon
2012 arXiv   pre-print
This work motivates the use of such "hard-to-optimize" non-submodular functionals, and proposes methods and algorithms to cope with the NP-hardness of their optimization.  ...  To overcome the resulting challenging optimization tasks the second axis of this thesis proposes approximation algorithms to cope with the NP-hardness of the optimization.  ...  We use ICM (Besag (1986) ), αβ-swap and α-expansion (large move making algorithms of Boykov et al. (2001) ) as representative single-scale "offthe-shelf" primal optimization algorithms.  ... 
arXiv:1210.7362v2 fatcat:6txhssbravatphb7tu6tedp6rm

A Survey on Label-efficient Deep Segmentation: Bridging the Gap between Weak Supervision and Dense Prediction [article]

Wei Shen, Zelin Peng, Xuehui Wang, Huayu Wang, Jiazhong Cen, Dongsheng Jiang, Lingxi Xie, Xiaokang Yang, Qi Tian
2022 arXiv   pre-print
To alleviate this burden, the past years have witnessed an increasing attention in building label-efficient, deep-learning-based segmentation algorithms.  ...  Next, we summarize the existing label-efficient segmentation methods from a unified perspective that discusses an important question: how to bridge the gap between weak supervision and dense prediction  ...  To address this issue, the authors introduced an adaptive affinity loss, where adaptive weights were adopted to measure the reliability of the pixel-wise affinities.  ... 
arXiv:2207.01223v1 fatcat:i7rgpxrfkrdbfm4effjdcjjr24

An experimental evaluation of the scalability of real-time scheduling algorithms on large-scale multicore platforms

Matthew Dellinger, Aaron Lindsay, Binoy Ravindran
2012 ACM Journal of Experimental Algorithmics  
We present an experimental analysis of the scalability of 16 multicore real-time scheduling algorithms. These algorithms include global, clustered, and partitioned algorithms.  ...  ChronOS extends these capabilities with a flexible, scalable real-time scheduling framework.  ...  A variety of schemes have been studied which utilize elements of both partitioning and global scheduling. The most common of these is known as clustered or semi-partitioned scheduling [21] .  ... 
doi:10.1145/2133803.2345677 fatcat:lev5h3a7qjaitj4ny2sd44zlli

ACOTES Project: Advanced Compiler Technologies for Embedded Streaming

Harm Munk, Eduard Ayguadé, Cédric Bastoul, Paul Carpenter, Zbigniew Chamski, Albert Cohen, Marco Cornero, Philippe Dumont, Marc Duranton, Mohammed Fellahi, Roger Ferrer, Razya Ladelsky (+15 others)
2010 International journal of parallel programming  
However, programming efficiently for streaming architectures is a challenging task, having to carefully partition the computation and map it to processes in a way that best matches the underlying streaming  ...  The analysis and transformation techniques automate large parts of the partitioning and mapping process, based on the properties of the application domain, on the quantitative information about the target  ...  It could be semi-static, where the compiler does not have to guess the dependencies -they are checked at run-time by the Hardware Task Scheduler.  ... 
doi:10.1007/s10766-010-0132-7 fatcat:dlvxlop65ngzfaezs3yjs2mfm4

Untangling cluster management with Helix

Kishore Gopalakrishna, Shi Lu, Zhen Zhang, Adam Silberstein, Kapil Surlaker, Ramesh Subramonian, Bob Schulman
2012 Proceedings of the Third ACM Symposium on Cloud Computing - SoCC '12  
These storage engines are robust building blocks used here for lowest-level storage, while other key features like partitioning, replication, etc. are built at higher layers.  ...  We describe the Helix design and implementation and present an experimental study that demonstrates its performance and functionality.  ...  An application running on Mesos consists of a scheduler and an executor. The scheduler registers with the master daemon to request resources.  ... 
doi:10.1145/2391229.2391248 dblp:conf/cloud/GopalakrishnaLZSSSS12 fatcat:2augvztpqfhdxkrlbif46qv2lu

Categorical Normalizing Flows via Continuous Transformations [article]

Phillip Lippe, Efstratios Gavves
2021 arXiv   pre-print
Using a factorized decoder, we introduce an inductive bias to model any interactions in the normalizing flow.  ...  By casting the encoding of categorical data in continuous space as a variational inference problem, we jointly optimize the continuous representation and the model likelihood.  ...  While being efficient in memory, they are slow in sampling and assume an order in the set of nodes.  ... 
arXiv:2006.09790v3 fatcat:dwecz4ikjzhppirfzay57xy2cq

Towards Single Stage Weakly Supervised Semantic Segmentation [article]

Peri Akiva, Kristin Dana
2021 arXiv   pre-print
We utilize point annotations to generate reliable, on-the-fly pseudo-masks through refined and filtered features.  ...  In contrary, our method offers a single-stage approach generalizable to arbitrary dataset, that is trainable from scratch, without any dependency on pre-trained backbones, classification, or separate refinement  ...  AffinityNet [3] lever- rectly thresholding such features will generate poor pseudo- ages image level labels to generate affinity labels obtained masks which will result in sub-optimal training  ... 
arXiv:2106.10309v2 fatcat:l3oafc7rz5frbiynru2vn6ogfa
« Previous Showing results 1 — 15 out of 961 results