industry news
Subscribe Now

Cavium Disrupts TCAM Market with Ultra Low Power NEURON Search™ Processor Family

San Jose, Calif., August 2, 2011— Cavium, Inc. (NASDAQ: CAVM), a leading provider of semiconductor products that enable intelligent processing for networking, communications, and the digital home, today announced its new game-changing NEURON Search™ Processor family targeted at a wide range of high performance, L2-L4 Network Search applications in Enterprise and Service Provider Infrastructure equipment. This new family includes the NEURON Search™ and NEURONMAX Search™ product lines with support for both IPv4 and IPv6 rules and delivers 100 million to over 1.6 billion searches per second with guaranteed low latency. The NEURON Search family delivers up to 4x the capacity per chip enabling the replacement of four existing state-of-the-art 40 Mbit TCAMs, while the NEURONMAX Search family enables virtually unlimited expansion at less than half the power consumption while dramatically slashing the cost, making them ideal for a wide range of Enterprise, Data Center and Wired/Wireless Service Provider applications. 

Discrete high-performance Search Processors and TCAMs are typically used for layer 2 to layer 4 lookups in higher-end Edge and Core Equipment where packet lookup requirements are greater than 50 million searches per second.  Two main trends in the networking and communication market are fueling the increased need for higher performance search capabilities. First, the transition from IPv4 addresses to the longer and much more complex IPv6 addresses. Second, the sheer explosion of traffic created by new mobile and social applications is driving 100Gbps port rates and also driving high-capacity line cards. According to the Cisco Visual Networking Index (VNI), by the end of 2015, annual global IP traffic will increase four-fold from 2010, video traffic will increase 32-fold and mobile traffic will increase 26-fold. The Linley Group forecasts that these drivers will push the market for search processor silicon to greater than $550 million in 2015. 

Expensive, power hungry legacy TCAMs and low performance algorithmic search processors until now have slowed the potential wider adoption of search technology. Cavium researchers took a completely different approach and developed breakthrough search technology using a range of mathematical and neural pattern recognition concepts that remove many of the critical constraints. Additionally, Cavium search technology sets the bar for silicon efficiency, improving up to 2x the number of rules that can be stored per unit area, leading to lower power and cost. This new cost power metric will drive wider adoption of search technology.

“Cavium is bringing breakthrough search technology to networking OEMs,” said Jag Bolaria, senior analyst at The Linley Group. “Compared with TCAMs currently in production, the NEURON Search Processor quadruples the capacity while the NEURONMAX Search family extends the capacity further. System designers will appreciate the lower power of these devices as well as having another established supplier of search processors.”

NEURON Search Processor Family Overview 

The NEURON Search Processor family consists of two silicon product lines that connect to a wide range of NPUs/ASICs including Cavium’s OCTEON II processors using a variety of serial and parallel IO interfaces such as the industry standard high bandwidth Interlaken/LA interface.

  • The NEURON Search Processor CNSP1XXX product line offers 100 million to 1.6 billion searches per second with enough capacity and guaranteed low latency in a single chip to replace four of the highest density 40 Mb TCAM parts available in the market today.  This product line is targeted at Enterprise, Data Center & Service Provider Infrastructure applications.
  • The NEURONMAX Search Processor CNSP2XXX product line enables seamless expansion of rule capacity through external DDR3 memory to 20x of the NEURON Search Processor family while retaining footprint compatibility, and is ideal for applications in target markets that require further future-proofing for a very large number of table entries.

Tremendous flexibility has been provided in storing rules and specifying rule table formats and a simple, easy to use SDK is provided for configuration, real-time updates and management.  The ease-of-use of this utility is in sharp contrast to the complex management software needed for TCAMs today and will help OEMs in rapidly integrating these products into their product lines. The software interface for all product families is identical thereby providing a highly scalable and software compatible solution.

Availability

Hardware design collateral including data sheet, pin out and Software Development Kit (SDK) are available now. SDK includes compiler, simulator and real-time rule update manager.  Both the NEURON Search and NEURONMAX Search product lines will be available in Q1 2012. Please contact your Cavium sales representative for additional details.

Leave a Reply

featured blogs
Feb 24, 2026
How a perfectly good Bosch HVAC system was undermined by preventable mistakes, and a thermostat interface that defies logic....

featured video

Cadence Chiplets Solutions | Helping you realize your chiplet ambitions

Sponsored by Cadence Design Systems

In this webinar, David Glasco, VP of Compute Solutions at Cadence, discusses how Cadence enables customers to transition from traditional monolithic SoC architectures to modular, scalable chiplet-based solutions, essential for meeting the growing demands of physical AI applications and high-performance computing.

Read eBook: Helping You Realize Your Chiplet Ambitions

featured chalk talk

The Han® Connector
Sponsored by Mouser Electronics and HARTING
In this episode of Chalk Talk, Emily Kenny from HARTING and Amelia Dalton investigate the details of the HARTING Han® connector family. They also explore the trends in connector solutions today, the variety of options within this connector family and how you can get started using a HARTING Han® connector for your next design!
Feb 18, 2026
9,786 views