Parallel Processing White Papers

Layout-Accurate Design and Implementation of a High-Throughput Interconnection Network for Single-Chip Parallel Processing

Overview A Mesh of Trees (MoT) on-chip interconnection network has been proposed recently to provide high throughput between memory units and processors for single-chip parallel processing. This paper reports the findings in bringing this concept to silicon. Specifically, it conducts cycle-accurate verilog simulations to verify the analytical results. The paper synthesizes and obtains the layout of the MoT interconnection networks of various sizes. To further improve throughput, the paper investigates different arbitration primitives to handle load and store, the two most common memory operations. The paper also studies the use of pipeline registers in large networks when there are long wires. Simulation based on full network layout demonstrates that significant throughput improvement can be achieved over the original proposed MoT interconnection network.

Further White Paper Details
PublisherUniversity of Maryland File FormatPDF
Date PublishedJune 2007
FormatWhite Papers   
Topics

Increasing Energy Efficiency with x86 Servers

Increasing energy efficiency with x86 servers. Robert Frances Group (RFG) explains why Intel/IBM is a winning combination.

Trend Watch: Mission Critcial x86

This paper examines how, over the past several years, research has shown many organisation moving mission-critical systems onto x86 servers.

Virtual SMB Centre in a Box?

What types of organizations have the most efficient IT operations? Data centers that provide highly available and resilient services with a minimum of human intervention and management? IT shops that can...

Virtualization strategy for mid-sized businesses

Virtualization of business applications allows IT operations in companies of all sizes to reduce costs, improve IT services and manage risk. The most dramatic cost savings are the result of reducing...

Get more from BI by understanding your analysis needs

Across your organization, different types of users - casual users, power users and executives - access data in order to answer different types of questions and then need to see...


Quick Sitemap Links: