You are here

Cypress to Present Advanced Classification Solutions at ACM SIGCOMM 2005 | Cypress Semiconductor

Cypress to Present Advanced Classification Solutions at ACM SIGCOMM 2005

Last Updated: November 24, 2011

Cypress to Present Advanced Classification Solutions at ACM SIGCOMM 2005

Application-specific, Algorithmic Solutions to Address Range Matching and Multiple-Entry Extraction in TCAMs

SAN JOSE, Calif., August 22, 2005 - Cypress Semiconductor Corp. (NYSE:CY) today announced that architects Dr. Srinivasan Venkatachary and Anand Rangarajan, along with UC Berkeley student Karthik Lakshminarayanan, will present their research into advanced packet classification using ternary content addressable memories (TCAMs) during a session at the SIGCOMM conference on August 24. They will present a method for using TCAMs to perform efficient range classification in packet processing systems, as well as a methodology for multiple-match extraction in TCAMs. The paper, Algorithms for Advanced Packet Classification with Ternary CAMs, can be found at

Range classification is the method by which a packet is classified by comparing a header field against a range of possible values to determine the policy that needs to be applied. This methodology is often used in network gateway systems to block access to certain types of incoming data, including unauthorized connections and other security threats. Multiple-match extraction techniques are designed to improve the flexibility of TCAM devices by providing the ability to return more than the first matching entry in the device. This is often required in load balancing and intrusion detection network applications.

"The range match and multiple match problems are coming to the forefront of packet classification problems due to the increased requirements for flexible classification and network protection," said Dr. Srinivasan Venkatachary. "Application-specific knowledge can be applied to create classification methods that are more optimal than traditional TCAM usage models allow. In the case of the range-matching problem, these methods can provide system vendors improved classification flexibility as well as more than double their current range matching capabilities."

About the Authors
Srinivasan Venkatachary is a principal architect at Cypress. He holds a PhD. in computer science from Washington University and was co-founder and Chief Scientist at Sahasra Networks, which Cypress acquired in 2002.

Anand Rangarajan is a systems architect and software engineering manager at Cypress. He holds a Masters of Science from Stanford University and was also a co-founder at Sahasra Networks.

Cypress's Network Search Engines
Cypress's search solution includes a comprehensive portfolio of products to address all segments of the search market.   In addition to its highly successful TCAM-based solutions (Ayama(TM) 10000 and Ayama 20000 NSEs), Cypress is the only vendor to also offer an algorithmic SRAM-based forwarding search engine (Sahasra(TM) 50000) to address the search needs of systems requiring large forwarding tables of more than 1 million table entries.  The Ayama 20000 NSEs are the first of their kind to offer a FastLink(TM) cascaded interface - a heterogeneous interface for optimized search subsystems which, when combined with Cypress's easy-to-use Cynapse(TM) software, enables seamless interworking of Cypress TCAM-based and algorithmic-based NSE families. Cypress's comprehensive NSE portfolio has been adopted by leading vendors in the router, L2/L3 switch, CMTS, DSLAM and storage area network (SAN) segments.

About Cypress
Cypress solutions are at the heart of any system that is built to perform: consumer, computation, data communications, automotive, industrial, and solar power. Leveraging a strong commitment to customer service and performance-based process and manufacturing expertise, Cypress's product portfolio includes a broad selection of wired and wireless USB devices, CMOS image sensors, timing solutions, network search engines, specialty memories, high-bandwidth synchronous and micropower memory products, optical solutions, and reconfigurable mixed-signal arrays. Cypress stock is traded on the New York Stock Exchange under the ticker symbol CY. More information about the company is available online at

# # #

Cypress and the Cypress logo are registered trademarks, and Ayama, Sahasra, FastLink and Cynapse are trademarks of Cypress Semiconductor Corporation. All other trademarks are the property of their respective owners.