Sciweavers

7891 search results - page 917 / 1579
» A Distributed Scheduling Framework
Sort
View
ICPP
1998
IEEE
15 years 11 months ago
Improving Parallel-Disk Buffer Management using Randomized Writeback
We address the problems of I/O scheduling and buffer management for general reference strings in a parallel I/O system. Using the standard parallel disk model withD disks and a sh...
Mahesh Kallahalla, Peter J. Varman
ICPP
1998
IEEE
15 years 11 months ago
A memory-layout oriented run-time technique for locality optimization
Exploiting locality at run-time is a complementary approach to a compiler approach for those applications with dynamic memory access patterns. This paper proposes a memory-layout ...
Yong Yan, Xiaodong Zhang, Zhao Zhang
IEEEPACT
1998
IEEE
15 years 11 months ago
Dynamic Hammock Predication for Non-Predicated Instruction Set Architectures
Conventional speculative architectures use branch prediction to evaluate the most likely execution path during program execution. However, certain branches are difficult to predic...
Artur Klauser, Todd M. Austin, Dirk Grunwald, Brad...
IPPS
1998
IEEE
15 years 11 months ago
Design and Implementation of a Parallel I/O Runtime System for Irregular Applications
In this paper we present the design, implementation and evaluation of a runtime system based on collective I/O techniques for irregular applications. We present two models, namely...
Jaechun No, Sung-Soon Park, Jesús Carretero...
HPCA
1997
IEEE
15 years 11 months ago
Datapath Design for a VLIW Video Signal Processor
This paper represents a design study of the datapath for a very long instruction word (VLIW) video signal processor (VSP). VLIW architectures provide high parallelism and excellen...
Andrew Wolfe, Jason Fritts, Santanu Dutta, Edil S....