Can I read A Reconfigurable CAM Architecture for Network Search Engines on EtoBox?

A Reconfigurable CAM Architecture for Network Search Engines by Mehrdad Nourani; Deepak S. Vijayasarathi; Poras T. Balsara is a scholarly article available to read on EtoBox.

What is A Reconfigurable CAM Architecture for Network Search Engines about?

A novel reconfigurable content addressable memory, called RCAM, is proposed that supports on-the-fly reconfiguration between CAM and TCAM. The area overhead of the proposed RCAM cell is only 5.6% when compared to conventional TCAM. This overhead is compensated by area saving due to removal of the priority encoder. Other features of our architecture include reconfigurability, and better overall performance and power. To achieve these we incorporated two novel techniques: (i) a hybrid CAM/TCAM architecture that allows user to pre-define CAM/TCAM cell behavior in each bit or word position and ultimately curtail the overall power consumptions of memory unit; and (ii) a wired-AND technique by which we can completely eliminate the sorting requirement and thus significantly reduce the update time. A 4Kb RCAM architecture was implemented using 0.18μm CMOS technology. The simulations indicate a search time of 6.15ns, i.e. capability of handling about five OC-192 at wire speed.

Author

Mehrdad Nourani; Deepak S. Vijayasarathi; Poras T. Balsara

Publisher

IEEE

Published

2006

Language

EN