Filters








2,091 Hits in 3.2 sec

Runtime Fragility in Main Memory [chapter]

Endre Palatinus, Jens Dittrich
2017 Lecture Notes in Computer Science  
Runtime fragility of the various data layouts for TPC-H queries Table 1 . 1 The schemas used in our experiments Label Schema Table 1 . 1 The schemas used in our experiments Label Schema Table 1 .  ...  vs. runtime layouts Experimental setup • memory resident tables (shm) • hand-coded queries implemented in C++ • single-threaded • process pinned to core • 3 warm-up runs, 5 warm runs Compiler + O−level  ... 
doi:10.1007/978-3-319-56111-0_9 fatcat:53xzyfupanf65e5ts3pbcpbmgm

Automatic speculative DOALL for clusters

Hanjun Kim, Nick P. Johnson, Jae W. Lee, Scott A. Mahlke, David I. August
2012 Proceedings of the Tenth International Symposium on Code Generation and Optimization - CHO '12  
Moreover, alias analysis is too fragile to achieve stable performance with small changes in the source code.  ...  the speculative runtime.  ...  Another issue with automatic parallelization is the fragility of static analysis. Figure 2 illustrates how fragile static analysis can be with a small change in the program.  ... 
doi:10.1145/2259016.2259029 dblp:conf/cgo/KimJLMA12 fatcat:rekpv7ckurahrafuvatz55o2ha

Principles of Antifragile Software

Martin Monperrus
2017 Proceedings of the International Conference on the Art, Science, and Engineering of Programming - Programming '17  
Automatic runtime bug fixing is changing the code in response to errors, fault injection in production means injecting errors in business critical software.  ...  Software may prevent errors by keeping a clean runtime envi-ronment, for instance by freeing the memory taken by disposed resources. Software detects errors with error detection code.  ...  Because of these three effects, injecting faults in production makes the system better. This corresponds to the main characteristic of antifragility: "the antifragile loves error".  ... 
doi:10.1145/3079368.3079412 dblp:conf/programming/Monperrus17 fatcat:arbhv7himnebjfqgt5pbxd5ssi

A comparison of knives for bread slicing

Alekh Jindal, Endre Palatinus, Vladimir Pavlov, Jens Dittrich
2013 Proceedings of the VLDB Endowment  
Vertical partitioning is a crucial step in physical database design in row-oriented databases.  ...  In fact, it is not even clear how to experimentally compare different vertical partitioning algorithms.  ...  access costs in main-memory.  ... 
doi:10.14778/2536336.2536338 fatcat:cw7ilrfhzzc6dmwkvwtjsxj3em

Locality-Aware Task Scheduling and Data Distribution for OpenMP Programs on NUMA Systems and Manycore Processors

Ananya Muddukrishna, Peter A. Jonsson, Mats Brorsson
2015 Scientific Programming  
However, techniques for distributing data are error-prone and fragile and require low-level architectural knowledge.  ...  Performance degradation due to nonuniform data access latencies has worsened on NUMA systems and can now be felt on-chip in manycore processors.  ...  Cache lines in a main memory page can be distributed uniformly to all home caches or to a single home cache.  ... 
doi:10.1155/2015/981759 fatcat:r6gun4qwdfbezg4nnx63zmo7nm

Reuse and Abuse

Susan Eisenbach, Chris Sadler
2007 Journal of Object Technology  
DYNAMIC LOADING AND LINKING The fragile base class problem arises in C++ because the C loader requires that its binaries are direct memory images.  ...  Limitations in memory and storage capacity led to the development of memory-resident (dynamic) link-libraries so that the application image size could be minimised and the library code shared between multiple  ... 
doi:10.5381/jot.2007.6.1.a5 fatcat:qhpx4hi5fja4vooeqdu5hjqxli

Core Working Set Based Scratchpad Memory Management

Ning DENG, Weixing JI, Jiaxin LI, Qi ZUO, Feng SHI
2011 IEICE transactions on information and systems  
Many state-of-the-art embedded systems adopt scratchpad memory (SPM) as the main on-chip memory due to its advantages in terms of energy consumption and on-chip area.  ...  The novelty of this SPM management method lies in its dependence on the program's dynamic access pattern as the main cue to conduct SPM allocation at runtime, thus offloading SPM management from the compiler  ...  Section 2 introduces the main idea of CWS and how it can impact SPM runtime allocation. Section 3 describes the CWS-based runtime SPM management in detail.  ... 
doi:10.1587/transinf.e94.d.274 fatcat:oh4naklq3nev3cxittg7cgahpq

Rozzle: De-cloaking Internet Malware

Clemens Kolbitsch, Benjamin Livshits, Benjamin Zorn, Christian Seifert
2012 2012 IEEE Symposium on Security and Privacy  
In addition, Rozzle triples the effectiveness of online runtime detection.  ...  While static and runtime methods for malware detection have been proposed in the literature, both on the client side, for just-intime in-browser detection, as well as offline, crawlerbased malware discovery  ...  Acknowledgments We appreciate the cooperation and efforts of David Felstead and Paul Rebriy of the Bing malware team in the course of this research.  ... 
doi:10.1109/sp.2012.48 dblp:conf/sp/KolbitschLZS12 fatcat:w7ciecc6jjfhzhsfjyilyw2jq4

Optimizing ROOT's Performance Using C++ Modules

Vassil Vassilev
2017 Journal of Physics, Conference Series  
Present state-of-art lazy parsing technique brings the runtime performance to reasonable levels but proves to be fragile and can introduce correctness issues.  ...  Exposing the full shared library descriptors to the interpreter at runtime translates into increased memory footprint.  ...  The recursive parsing mode is fragile and error-prone because it is not a welldefined language feature.  ... 
doi:10.1088/1742-6596/898/7/072023 fatcat:o2qguuytjfa3hm67n5kvqbyd4e

Topology-Aware and Dependence-Aware Scheduling and Memory Allocation for Task-Parallel Languages

Andi Drebes, Karine Heydemann, Nathalie Drach, Antoniu Pop, Albert Cohen
2014 ACM Transactions on Architecture and Code Optimization (TACO)  
Experimental results, based on the Open-Stream language, show that locality of accesses to main memory of scientific applications can be increased significantly on a 64-core machine, resulting in a speedup  ...  Task and data placement decisions are based on a static description of the memory hierarchy and on runtime information about intertask communication.  ...  to main memory.  ... 
doi:10.1145/2641764 fatcat:tlu32qcxpnbsxpgikiw5blbg5e

Temporality a NVRAM-based Virtualization Platform

Vasily A. Sartakov, Arthur Martens, Ruediger Kapitza
2015 2015 IEEE 34th Symposium on Reliable Distributed Systems (SRDS)  
Novel persistent main memory, i.e. memory that retains stored data without an external source of power, firstly prevents data loss in case of a power outage, secondly reduces the time for a system reboot  ...  Yet new architectures and programming models are required to utilize persistent main memory.  ...  CPU, DRAM main memory and hard drives.  ... 
doi:10.1109/srds.2015.42 dblp:conf/srds/SartakovMK15 fatcat:lyu4isj7ejfwvbmgr4lgjnchue

Paragon

Mehrzad Samadi, Amir Hormati, Janghaeng Lee, Scott Mahlke
2012 Proceedings of the 5th Annual Workshop on General Purpose Processing with Graphics Processing Units - GPGPU-5  
The rise of graphics engines as one of the main parallel platforms for general purpose computing has ignited a wide search for better programming support for GPUs.  ...  In this work, we take a different approach.  ...  The main problem with this approach is that the compiler analyses used for auto-parallelization are usually too conservative and fragile resulting in very small or no gains.  ... 
doi:10.1145/2159430.2159438 dblp:conf/asplos/SamadiHLM12 fatcat:ushyisgje5a3haujbwsrridyxu

Dynamic and Speculative Polyhedral Parallelization Using Compiler-Generated Skeletons

Alexandra Jimborean, Philippe Clauss, Jean-François Dollinger, Vincent Loechner, Juan Manuel Martinez Caamaño
2013 International journal of parallel programming  
The framework has been implemented with extensions of the LLVM compiler and an x86-64 runtime system.  ...  During this phase, the accessed memory addresses are used to compute on-the-fly dependence distance vectors, and are also interpolated to build a predictor of the forthcoming accesses.  ...  The main difference is that our skeletons (also generated at compile-time), still contain parameterized code which is instantiated at runtime.  ... 
doi:10.1007/s10766-013-0259-4 fatcat:efm2hzveyngsfpbv24rundkuom

Block-Parallel IDA* for GPUs (Extended Manuscript) [article]

Satoru Horie, Alex Fukunaga
2017 arXiv   pre-print
We propose Block-Parallel IDA* (BPIDA*), which assigns the search of a subtree to a block (a group of threads with access to fast shared memory) rather than a thread.  ...  Background and Related Work An NVIDIA CUDA architecture GPU consists of a set of streaming multiprocessors (SMs) and a GPU main memory (shared among all SMs).  ...  Thus, all work could be performed in the SM local+shared memories, without using global memory.  ... 
arXiv:1705.02843v1 fatcat:tie5kpiwpvdqzjhecoocmkqgea

TripleAgent: Monitoring, Perturbation and Failure-obliviousness for Automated Resilience Improvement in Java Applications [article]

Long Zhang, Martin Monperrus
2019 arXiv   pre-print
In this paper, we present a novel resilience improvement system for Java applications.  ...  The latter is achieved thanks to the failure-oblivious computing, a concept introduced in 2004 by Rinard and colleagues. We design and implement the system as agents for the Java virtual machine.  ...  In Java, there is no invalid memory addresses, but the biggest reason for crashing are exceptions thrown at runtime. Thus, we do failureoblivious computing for uncaught and mishandled exceptions.  ... 
arXiv:1812.10706v3 fatcat:vzbbo5jvzjba3pikjkyeb5o7pi
« Previous Showing results 1 — 15 out of 2,091 results