1.41 ‘Giga-Searches’ per Second?
By: Igor Arsovski
GLOBALFOUNDRIES’ Ternary Content Addressable Memory Core sets a new performance and density record enabling faster cloud and data center traffic.
Data centers, which are the engines of the Internet, are responsible for the vast amount of traffic that flows across the network. And, it’s no revelation; our digital universe and data center traffic will reach 10.4 zettabytes (ZB) by 2019, the equivalent to 144 trillion hours of streaming music. As apps and services become more data hungry, the higher the allowance for data traffic we’re going to need to feed them.
And, getting there will require innovative solutions – to improve scalability and achieve higher throughput with lower latency, data centers use Ternary Content Addressable memory (TCAM). TCAM, an application specific memory, is designed to help data centers speed the search of large look-up tables.
GF recently introduced a new ASIC search architecture TCAM solution at the International Solid-State Circuits Conference in San Francisco. GF’s TCAM core demonstrated the highest published search rate of 1.4 billion searches per second core, while also achieving a record density of 2Mb/mm2. This is good news for data centers, which commonly use TCAM in applications such as pattern recognition, data compression, network security and packet forwarding.
Furthermore, GF’s TCAM design uses both proprietary architecture and circuits for power supply pre-conditioning that reduce power supply noise collapse by 50 percent, enabling a path to further performance and density scaling. These improvements are critical to enable faster data centers with larger network tables capable of handling the constantly increasing network traffic.
TCAMs are extremely complex memory designs, which is why GF’s expertise – more than 20 years of pushing the limits of TCAM design – has enabled us to deliver improved search throughput, high density, and low power consumption for our customers’ ASIC designs. Specifically, it may be our seven generations of embedded TCAM design, qualification and debug experience, with more than 50 ASICs with large TCAM content in production, that allowed us to provide the highest search-bandwidth (fastest) networking TCAM in the industry.
To learn more about GF’s ASIC solutions, download our ASIC FX-14 technology brief or contact a GF ASIC sales representative.
Fellow, Essex Junction, VT
Igor Arsovski is a GLOBALFOUNDRIES Fellow in the ASIC IP Design Group in Essex Junction, Vermont. He is responsible for defining power performance and area targets for memory IP and leads TCAM architecture and circuit definition including high-performance SRAM design. His extended focus is energy efficient building blocks for Deep Learning, High-Reliability Automotive Electronics, and 2.5/3D Memory Integration.
Prior to joining GF in 2015, Igor was senior technical staff member in the IBM Microelectronics group.
Igor holds a Master’s degree from the University of Toronto. He is an author of 14 IEEE papers, and has over 80 issued or pending patents.