Filters








153,663 Hits in 9.8 sec

A case for user-level dynamic page migration

Dimitrios S. Nikolopoulos, Theodore S. Papatheodorou, Constantine D. Polychronopoulos, Jesús Labarta, Eduard Ayguadé
2000 Proceedings of the 14th international conference on Supercomputing - ICS '00  
This paper presents user-level dynamic page migration, a runtime technique which transparently enables parallel programs to tune their memory performance on distributed shared memory multiprocessors, with  ...  We also present a new technique for preventing page pingpong and a mechanism for monitoring the performance of page migration algorithms at runtime and tuning their sensitive parameters accordingly.  ...  As a consequence, the use of dynamic page migration in contemporary operating systems for scalable NUMA multiprocessors is either discouraged or limited to trivial cases, such as the replication of kernel  ... 
doi:10.1145/335231.335243 dblp:conf/ics/NikolopoulosPPLA00 fatcat:cc4ck37on5fdvdygx2f54mpn64

Is Data Distribution Necessary in OpenMP?

D.S. Nikolopoulos, T.S. Papatheodorou, C.D. Polychronopoulos, J. Labarta, E. Ayguade
2000 ACM/IEEE SC 2000 Conference (SC'00)  
We also show that performance leaks stemming from suboptimal page placement schemes can be remedied with a smart user-level page migration engine.  ...  The main body of the paper describes how the OpenMP runtime environment can use page migration for implementing implicit data distribution and redistribution schemes without programmer intervention.  ...  User-level dynamic page migration To investigate the issues stated before, we have developed a runtime system called UPMlib (user-level page migration library), which injects a dynamic page migration engine  ... 
doi:10.1109/sc.2000.10025 dblp:conf/sc/NikolopoulosPPLA00 fatcat:rtv6tp4gofgitclqgk3dennrtm

A Transparent Runtime Data Distribution Engine for OpenMP

Dimitrios S. Nikolopoulos, Theodore S. Papatheodorou, Constantine D. Polychronopoulos, Jesús Labarta, Eduard Ayguadé
2000 Scientific Programming  
Second, the paper presents a transparent, user-level page migration engine with an ability to gain back any performance loss that stems from suboptimal placement of pages in iterative OpenMP programs.  ...  The main body of the paper describes how our OpenMP runtime environment uses page migration for implementing implicit data distribution and redistribution schemes without programmer intervention.  ...  The same work proposed a complete kernel-level implementation of a dynamic page migration engine and evaluated it using accurate machine-level simulation.  ... 
doi:10.1155/2000/417570 fatcat:54ksufgvujh5tpxrxlfoigiclu

Leveraging Transparent Data Distribution in OpenMP via User-Level Dynamic Page Migration [chapter]

Dimitrios S. Nikolopoulos, Theodore S. Papatheodorou, Constantine D. Polychronopoulos, Jesús Labarta, Eduard Ayguadé
2000 Lecture Notes in Computer Science  
We have implemented a runtime system called UPMlib, which allows the compiler to inject into the application a smart user-level page migration engine.  ...  The vehicle for implementing these facilities in OpenMP without modifying the programming model or exporting data distribution details to the programmer is user-level dynamic page migration [9,10].  ...  dynamic page migration at user-level.  ... 
doi:10.1007/3-540-39999-2_40 fatcat:e7tdiomq3nc6bgemkkuufaacfi

Evaluation of the memory page migration influence in the system performance

Julita Corbalan, Xavier Martorell, Jesus Labarta
2003 Proceedings of the 17th annual international conference on Supercomputing - ICS '03  
To reduce the impact of a bad initial memory placement, some operating systems offer a dynamic memory migration mechanism.  ...  In this paper, we want to demonstrate that memory migration mechanisms are a useful approach, but that their performance depends more on related issues, such as the processor scheduling, than on the mechanism  ...  Although this execution environment is so hard for the applications to obtain performance, the authors show that, in this case, their proposal for dynamic page migrations implemented from user-level provides  ... 
doi:10.1145/782814.782833 dblp:conf/ics/CorbalanML03 fatcat:t6bjbcuygvh33pm5y4xtkmplly

Evaluation of the memory page migration influence in the system performance

Julita Corbalan, Xavier Martorell, Jesus Labarta
2003 Proceedings of the 17th annual international conference on Supercomputing - ICS '03  
To reduce the impact of a bad initial memory placement, some operating systems offer a dynamic memory migration mechanism.  ...  In this paper, we want to demonstrate that memory migration mechanisms are a useful approach, but that their performance depends more on related issues, such as the processor scheduling, than on the mechanism  ...  Although this execution environment is so hard for the applications to obtain performance, the authors show that, in this case, their proposal for dynamic page migrations implemented from user-level provides  ... 
doi:10.1145/782832.782833 fatcat:2iss3j5ytja4dpifvm3u47zp7a

Scheduler-Activated Dynamic Page Migration for Multiprogrammed DSM Multiprocessors

Dimitrios S. Nikolopoulos, Constantine D. Polychronopoulos, Theodore S. Papatheodorou, Jesús Labarta, Eduard Ayguadé
2002 Journal of Parallel and Distributed Computing  
This paper presents a novel dynamic page migration algorithm that remedies this problem in iterative parallel programs.  ...  The key mechanism that enables timely identification of these pages is a communication interface between the page migration engine and the operating system scheduler.  ...  ACKNOWLEDGMENT We thank the journal referees for their insightful comments, which helped us improve the paper considerably.  ... 
doi:10.1006/jpdc.2001.1817 fatcat:mm4g6niwc5e4dn4adqub77grbu

Geographical Locality and Dynamic Data Migration for OpenMP Implementations of Adaptive PDE Solvers [chapter]

Markus Nordén, Henrik Löf, Jarmo Rantakokko, Sverker Holmgren
2008 Lecture Notes in Computer Science  
The main conclusions of the study are: (1) that geographical locality is very important for the performance of the solver, (2) that the performance can be improved significantly using dynamic page migration  ...  of misplaced data, (3) that a migrate-on-next-touch directive works well whereas the first-touch strategy is less advantageous for programs exhibiting a dynamically changing memory access patterns, and  ...  In Solaris, dynamic migration of pages between nodes can be performed using a directive with a migrate-on-next-touch semantic using the madvise(3C) library call [20] .  ... 
doi:10.1007/978-3-540-68555-5_31 fatcat:5grtv5jzrzdkfpzbg3fghhitha

Sentinel: Runtime Data Management on Heterogeneous Main MemorySystems for Deep Learning [article]

Jie Ren, Jiaolin Luo, Kai Wu, Minjia Zhang, Dong Li
2019 arXiv   pre-print
To achieve this,Sentinel exploits domain knowledge about deep learning to adopt a custom approach for data management.  ...  Sentinel leverages workload repeatability to break the dilemma between profiling accuracy and overhead; It enables profiling and data migration at the granularity of data objects (not pages), by controlling  ...  Using this method, page-level profiling becomes data object-level profiling.  ... 
arXiv:1909.05182v1 fatcat:vh27yunrjndo7nzp34aqdy2xnu

Memos: Revisiting Hybrid Memory Management in Modern Operating System [article]

Lei Liu, Mengyao Xie, Hao Yang
2017 arXiv   pre-print
Powered by our newly designed kernel-level monitoring module and page migration engine, memos can dynamically optimize the data placement at the memory hierarchy in terms of the on-line memory patterns  ...  Generally, Fig.1 tells us, for a specific page, the WD/RD and Hot/Cold patterns may change during run time, and thus a dynamic mechanism is needed to migrate data across NVM and DRAM.  ...  In case 3, memos migrates a hot/WD page from NVM to DRAM across channels.  ... 
arXiv:1703.07725v1 fatcat:wzdseqs4pbc7nozfejrllqszl4

Migrating Multi-page Web Applications to Single-page AJAX Interfaces

Ali Mesbah, Arie van Deursen
2007 11th European Conference on Software Maintenance and Reengineering (CSMR'07)  
Our approach, using a schema-based clustering technique, extracts a navigational model of web applications, and identifies candidate user interface components to be migrated to a single-page AJAX interface  ...  If until a year ago, the concern revolved around migrating legacy systems to web-based settings, today we have a new challenge of migrating web applications to single-page AJAX applications.  ...  Acknowledgments This work received partial support from Sen-terNovem, project Single Page Computer Interaction (SPCI). We thank the anonymous referees for their valuable comments.  ... 
doi:10.1109/csmr.2007.33 dblp:conf/csmr/MesbahD07 fatcat:tirmwzs7rrgjpcasgzxbldptpq

Data placement in HPC architectures with heterogeneous off-chip memory

Milan Pavlovic, Nikola Puzovic, Alex Ramirez
2013 2013 IEEE 31st International Conference on Computer Design (ICCD)  
Our results show that the hybrid memory system with dynamic page migration and limited DRAM capacity, can achieve performance that is comparable to a hypothetical, hard to implement, DRAM-only system.  ...  In this paper, we propose an architecture with a hybrid memory design that places two technologically different memory modules in a flat address space.  ...  Dynamic page migration Finally, a Dynamic policy introduces page migration in the other direction (from PCM to DRAM), denoted as backmigration.  ... 
doi:10.1109/iccd.2013.6657042 dblp:conf/iccd/PavlovicPR13 fatcat:ruw362p5pndu5jrjol5qojdicy

Understanding the Behavior of Pthread Applications on Non-Uniform Cache Architectures

Gagandeep S. Sachdev, Kshitij Sudan, Mary W. Hall, Rajeev Balasubramonian
2011 2011 International Conference on Parallel Architectures and Compilation Techniques  
In this paper, we take existing parallel programs written in Pthreads, and show the performance gap between current static mapping schemes, costly migration schemes and idealized static and dynamic best-case  ...  Future scalable multi-core chips are expected to implement a shared last-level cache (LLC) with banks distributed on chip, forcing a core to incur non-uniform access latencies to each bank.  ...  We intend to pursue the following optimizations to improve the performance of the code using a softwareonly strategy: (1) using thread-private storage when safe for stack and local, dynamically-allocated  ... 
doi:10.1109/pact.2011.26 dblp:conf/IEEEpact/SachdevSHB11 fatcat:oukhzso76vc7noqenro5umr2p4

Improving Data Locality Using Dynamic Page Migration Based on Memory Access Histograms [chapter]

Jie Tao, Martin Schulz, Wolfgang Karl
2002 Lecture Notes in Computer Science  
The approach is based on run-time data redistribution via user-level page migration.  ...  First performance experiments on several applications show the potential for a significant gain in speedup.  ...  Improving Data Locality Using Dynamic Page Migration tion tool, a deep insight into the memory access patterns of the application and thereby enables further optimizations.  ... 
doi:10.1007/3-540-46080-2_98 fatcat:kgaydtgd55fhto2nwbscakfb5y

An Operating System Level Data Migration Scheme in Hybrid DRAM-NVM Memory Architecture [article]

Reza Salkhordeh, Hossein Asadi
2018 arXiv   pre-print
In the proposed scheme, two Least Recently Used (LRU) queues, one for DRAM section and one for NVM section, are used for the sake of data migration.  ...  This paper presents an efficient data migration scheme at the Operating System level in a hybrid DRAMNVM memory architecture.  ...  In addition to the traditional moving pages in case of a miss or evicting a data page, hybrid memories have migrations between two memories.  ... 
arXiv:1805.02514v1 fatcat:s3xdo5xpqrgclnnfaqhpvwj5fm
« Previous Showing results 1 — 15 out of 153,663 results