Search results
-
Achieving 10Gbps Line-rate Key-value Stores with FPGAs
application. In this paper we present the design of a novel memcached architecture implemented on Field ...michele - January 25, 2022 - 12:38 pm
-
Using Set Cover to Optimize a Large-Scale Low Latency Distributed Graph
from many machines. In this paper, we present a modified greedy set cover algorithm that is used to ...michele - January 25, 2022 - 2:03 pm
-
Dissecting Open Source Cloud Evolution: An OpenStack Case Study
Salman A. Baset, Chunqiang Tang, Byung Chul Tak, and Long Wang, IBM T.J. Watson Research Center Open source cloud platforms are playing an increasingly significant role in cloud computing. These systems have been undergoing rapid development cycles. As an ...michele - January 25, 2022 - 11:13 am
-
The Case for Limping-Hardware Tolerant Clouds
issues such as hardware failures, software bugs, and configuration mistakes. In this paper, we highlight one ...michele - January 25, 2022 - 12:51 pm
-
Cloud Computing for the Power Grid: From Service Composition to Assured Clouds
on communications and computation to plan, operate and analyze power systems. In this paper we explore the reasons ...michele - January 25, 2022 - 12:45 pm
-
Towards a Fault-Resilient Cloud Management Stack
Xiaoen Ju, University of Michigan; Livio Soares, IBM T.J. Watson Research Center; Kang G. Shin, University of Michigan; Kyung Dong Ryu, IBM T.J. Watson Research Center Cloud management stacks have become a new important layer in cloud computing infras ...michele - January 25, 2022 - 1:05 pm
-
Parallelization Primitives for Dynamic Sparse Computations
Tsung-Han Lin, Stephen J. Tarsa, and H.T. Kung, Harvard University We characterize a general class of algorithms common in machine learning, scientific computing, and signal processing, whose computational dependencies are both sparse, and dynamically def ...michele - December 15, 2021 - 10:52 pm
-
Three Fingered Jack: Tackling Portability, Performance, and Productivity with Auto-Parallelized Python
David Sheffield, Michael Anderson, and Kurt Keutzer, University of California, Berkeley We present Three Fingered Jack, a highly productive approach to writing high-performance software in Python. Our system applies traditional dependence analysis and reo ...michele - December 15, 2021 - 10:52 pm
-
Collection-focused Parallelism
Micah J. Best and Daniel Jacobsen, University of British Columbia; Nicholas Vining, Gaslamp Games; Alexandra Fedorova, Simon Fraser University Constructing parallel software is, in essence, the process of associating ‘work’ with computational units. Th ...michele - December 15, 2021 - 10:52 pm
-
Constrained Data-Driven Parallelism
Tim Harris, Yossi Lev, Victor Luchangco, Virendra J. Marathe, and Mark Moir, Oracle Labs In data-driven parallelism, changes to data spawn new tasks, which may change more data, spawning yet more tasks. Computation propagates until no further changes occu ...michele - December 15, 2021 - 10:52 pm
-
Towards Performance-Portable, Scalable, and Convenient Linear Algebra
Philippe Tillet, Technische Universität Wien; Karl Rupp, Argonne National Laboratory; Siegfried Selberherr, Technische Universität Wien; Chin-Teng Lin, National Chiao Tung University The rise of multi- and many-core architectures also gave birth to ...michele - December 15, 2021 - 10:52 pm
-
Determinism Is Overrated: What Really Makes Multithreaded Programs Hard to Get Right and What Can Be Done About It
multithreading. In this paper, we argue that determinism is not as useful as commonly perceived: it is neither ...michele - December 15, 2021 - 10:52 pm
-
…And Region Serializability for All
a difficult trade-off between losing all semantic guarantees and hurting performance. In this paper, we argue ...michele - December 15, 2021 - 10:52 pm
-
Durability Semantics for Lock-based Multithreaded Programs
especially in a multithreaded programming environment. In this paper, we provide semantics for identifying ...michele - December 15, 2021 - 10:52 pm
-
Characterizing Real World Bugs Causing Sequential Consistency Violations
incorrectness. This paper provides the first (to the best of our knowledge) comprehensive characteristics study ...michele - December 15, 2021 - 11:52 pm
-
But How Do We Really Debug Transactional Memory Programs?
utilize unrealistic hardware. In this paper, we motivate the need for more than traditional and ad hoc ...michele - December 15, 2021 - 11:52 pm
-
Property-Driven Cooperative Logging for Concurrency Bugs Replication
Técnica de Lisboa This paper presents property-driven partial logging, a novel information partition ...michele - December 15, 2021 - 11:52 pm
-
Parallel Synchronization-Free Approximate Data Structure Construction
Martin Rinard, MIT CSAIL We present approximate data structures with construction algorithms that execute without synchronization. The data races present in these algorithms may cause them to drop inserted or appended elements. Nevertheless, the algorithm ...michele - December 15, 2021 - 11:52 pm
-
Using Elimination and Delegation to Implement a Scalable NUMA-Friendly Stack
Irina Calciu, Brown University; Justin Gottschlich, Intel Labs; Maurice Herlihy, Brown University Emerging cache-coherent non-uniform memory access (cc-NUMA) architectures provide cache coherence across hundreds of cores. These architectures change how ...michele - December 15, 2021 - 11:52 pm
-
Modeling Mobile Cloud Computing Using Greenmetrics
cloud resources. This paper tackles a suite of optimization subproblems: a program partitioning, network ...michele - December 15, 2021 - 8:52 pm
-
On the Stability of a Market-Oriented Cloud Computing Model with Time-Varying Workloads
José Marcio Luna and C.T. Abdallah, and G.L. Heileman, University of New Mexico In this paper we ...michele - December 15, 2021 - 8:52 pm
-
Online Resource Management for Data Center with Energy Capping
A. S. M. Hasan Mahmud and Shaolei Ren, Florida International University Awarded Best Paper! The ... paper, we study energy budgeting and propose a novel online resource management algorithm, called ORM, to ...michele - December 15, 2021 - 8:52 pm
-
Adaptive Control of Apache Web Server
varying traffic. To address this problem, we design in this paper an adaptive controller for the Apache ...michele - December 15, 2021 - 8:52 pm
-
Two-level Throughput and Latency IO Control for Parallel File Systems
high-performance computing (HPC) systems continue to scale up and be used in a shared environment. This paper ...michele - December 15, 2021 - 8:52 pm
-
Towards Model-based Management of Database Fragmentation
Asim Ali, Qatar University; Rui Jia, Mississippi State University; Abdelkarim Erradi, Qatar University; Sherif Abdelwahed, Mississippi State University; Rachid Hadjidj, Qatar University The performance of a database can significantly deteriorate due to th ...michele - December 15, 2021 - 8:52 pm