Study & Performance Comparison of Positive Feedback Match-Line (ML) Sensing Schemes for Low Power TCAM

Abstract

Ternary content addressable memories (TCAMs) are hardware-based parallel lookup tables with bit-level masking capability. They are attractive for applications such as packet forwarding and classification in network routers especially in internet applications. Despite the attractive features of TCAMs, high power consumption is one of the most critical drawbacks of TCAM. Hence reducing the power consumption without sacrificing the speed and voltage margin is the most difficult part in TCAM design. Among different match line sensing schemes, the use of positive feedback in the sense amplifiers is one of the best solutions to this problem. The main feather of this work is to perform comparison among different existing positive feedback based match line sensing schemes, i.e., mismatch dependent, active feedback and resistive feedback schemes using four performance parameters which are (i) search time (ii) voltage margin (iii) peak dynamic power and (iv)worst case energy consumption. All the schemes are simulated using 130nm, 1.2V CMOS logic. It is shown in this work that the energy saving is maximum (68.77%) in resistive feedback scheme compared to conventional CR-MLSA. Again, comparing among the positive feedback based schemes it is found that the resistive feedback provides with the best speed. Mismatch dependent scheme provides the best voltage margin and peak dynamic power. The worst case energy consumption is least in Active feedback scheme among all three positive feedback based scheme.