| Publisher | SGI | ||
|---|---|---|---|
| Format | 135.6KB PDF, requires Acrobat Rdr 5 | Date added | 30 Apr 2003 |
| Topics | Supercomputers, High Performance Computing | ||
| Downloads | 827 | ||
This paper describes the global shared memory architecture and benefits of the SGI Altix 3000 family of servers and super-clusters. Memory size and scaling, cache organization and coherency, nodes and inter-node memory access, bandwidths and latencies, and RAS features are all discussed. The paper is written from a perspective that may be useful to an applications developer or a system administrator. It describes the current implementation of the Altix 3000 system communication infrastructure.
Did you find this white paper useful?
39 out of 50 users found this white paper useful
Related white papers
Supercomputing Center Sees 46 Percent Performance Gain With HPC System
The Las Vegas, Nevada - based National Supercomputing Center for Energy and the Environment is a full-service supercomputing facility with a mission for excellence in education and research. The National...
Sun Microsystems' Processor and Network Products Group: Sun Ray Ultra-Thin Client Success Story
Electronic design automation (EDA) is the epitome of compute-intensive applications. The hallmark of EDA computing is performing incredibly high numbers of computations on unbelievably large amounts of data. The mission...
University Improves Research and Saves 30 Per Cent of Budget With a High-Performance Computing Cluster
The department of physics at University Alexandru Ioan Cuza of Iasi was looking for a cost-effective way to build a supercomputer that was powerful enough to run complex mathematical research...
The Engineer Research and Development Center Leverages HP Supercomputer Systems to Help the Department of Defense Support Soldiers in Combat
U.S. Army Engineer Research and Development Center ERDC) Major Shared Resource Center (MSRC), managed as part of the Department of Defense-wide high performance computing capability provided by the DoD High...
Application Sensitivity to Link and Injection Bandwidth on a Cray XT4 System
This paper describes the efforts to characterize application sensitivity to link and injection bandwidth on a Cray XT4 system. Link bandwidth is controlled by modifying the number of rails activated...
Leading University Creates UK's Fastest Academic Supercomputer Through Collaborative Partnership
Cambridge University has had supercomputers for use in complex research since 1992. Undertaking ground breaking research, users require cutting edge technology that provides optimum performance. At the end of a...
IBM Virtualization Manager Demo
IBM Virtualization Manager allows you to discover, visualize, and manage both physical and virtual systems from a single console. View the demo to learn more.
