|
One billion searches per second (BSPS) performance
|
Supports multiple lookups for packet processors providing aggregate 10G (10GbE, 8x1GbE, OC192) and 40G (4x10GbE, OC-768) performance
|
|
Quad Simultaneous Multi-Database Lookup (qSMDL)
|
Allows our customers to search multiple application databases with the same search key, allowing simultaneous comparisons of IP forwarding, Netflow, access control, and QoS/CoS databases to reduce system latency and improve overall system line-rate performance
|
|
Base search width of 80-bits
|
Only architecture truly optimized for IPv6 by offering a 320-bit search width to accommodate typical 296-bit wide IPv6 5-tuple look-ups, improving search performance to accelerate application delivery for IPv6-enabled networks as well as enhancing storage capacity for IPv6 entries
|
|
Double Data Rate (DDR) interface with equivalent clock input and core search rate
|
Simplified clocking scheme enables designers to use less expensive components with improved signal integrity and a simplified termination scheme, decreasing the cost and complexity of the overall design
|
|
Integrated error code correction and bus parity scan engine
|
By improving data integrity and addressing any soft errors that may be induced by the 1V core voltage, helps our customers improve system reliability and guarantee higher system uptime to their service provider and enterprise customers
|
|
Dynamic Database Management and Iterative Entry Management
|
Two power saving techniques that work in tandem to selectively reduce the number of entries being searched during table lookups to reduce power consumption by up to 75%
|
|
Automatic Entry Recycling
|
Tracks and reports entry usage within the network search accelerator, identifying entries that are infrequently used for easy removal and replacement
|