Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network

  Advanced Search  

Support for High-Frequency Streaming in CMPs Ram Rangan Neil Vachharajani Adam Stoler

Summary: Support for High-Frequency Streaming in CMPs
Ram Rangan Neil Vachharajani Adam Stoler
Guilherme Ottoni David I. August George Z. N. Cai
Departments of Computer Science and Electrical Engineering Digital Enterprise Group
Princeton University Intel Corporation
Princeton NJ 08540 Hillsboro OR 97124
{ram, nvachhar, astoler, ottoni, august}@princeton.edu george.cai@intel.com
As the industry moves toward larger-scale chip multiprocessors, the
need to parallelize applications grows. High inter-thread commu-
nication delays, exacerbated by over-stressed high-latency memory
subsystems and ever-increasing wire delays, require parallelization
techniques to create partially or fully independent threads to im-
prove performance. Unfortunately, developers and compilers alike
often fail to find sufficient independent work of this kind.
Recently proposed pipelined streaming techniques have shown
significant promise for both manual and automatic parallelization.
These techniques have wide-scale applicability because they em-
brace inter-thread dependences (albeit acyclic dependences) and
tolerate long-latency communication of these dependences. This


Source: August, David - Department of Computer Science, Princeton University


Collections: Computer Technologies and Information Sciences