Architectural choices for the Columbia 0.8 Teraflops machine

Preprint

20 December 1994

preprint
Published by arXiv in arXiv

https://doi.org/10.48550/arXiv.hep-lat/9412093

Abstract

We discuss the hardware design choices made in our 16K-node 0.8 Teraflops supercomputer project, a machine architecture optimized for full QCD calculations. The efficiency of the conjugate gradient algorithm in terms of balance of floating-point operations, memory handling and utilization, and communication overhead is addressed. We also discuss the technological innovations and software tools that facilitate hardware design and what opportunities these give to the academic community.

Keywords

TECHNOLOGICAL INNOVATIONS
ARCHITECTURAL CHOICES
TERAFLOPS MACHINE
HARDWARE DESIGN
CHOICES MADE
ARCHITECTURE
EFFICIENCY
ALGORITHM
QCD

All Related Versions

Version 1, 1994-12-20, ArXiv
Published version: Nuclear Physics B - Proceedings Supplements, 42 (1-3), 902.

This publication has 0 references indexed in Scilit: