Broadcom Corporation (NASDAQ: BRCM), a global innovation leader in semiconductor solutions for wired and wireless communications, today announced the industry's first 100 Gbps full duplex network processor unit (NPU). Enabling the next wave of 100GbE optimized switches and routers for service provider networks, the fully programmable BCM88030 family features 64 custom processors running at 1GHz, delivering more than 2X the throughput of any NPU on the market. With the industry's highest level of integration, the BCM88030 solution eliminates costly external components, dramatically reducing system cost and power by up to 80 percent per 10 GbE port.

In 2012, the number of connected devices is set to exceed the world's population with an estimated 7 billion devices connected to the network(1). By 2015, 90 percent of content viewed on mobile devices will be streaming video and application downloads are expected to reach 47 billion per year(1). To satisfy this overwhelming appetite for bandwidth, service providers around the globe are racing to transform their networks by adopting higher bandwidth links. Analysts estimate a compound annual growth rate (CAGR) of more than 170 percent from 2011 to 2016 for 100G Ethernet ports as service providers rush to meet this demand(2).

Built on the proven XGS Core packet processing architecture, the BCM88030 family delivers the industry's highest performance NPU in a single device. With a powerful array of high-speed processors, extensive multi-threading and hardware acceleration for functions such as packet parsing, classification and look-ups, the BCM88030 product family delivers a feature-rich application at 100GbE full duplex for complex service provider switching and routing requirements. With a robust software development environment, the BCM88030 family is completely user programmable, enabling a highly flexible forwarding solution that ensures support for the most demanding applications, both existing and future.

The BCM88030 NPU family includes a proprietary algorithmic look up engine using low cost DDR-3 DRAM that enables massive scale for Layer 2, IPv4 and IPv6 tables while significantly reducing system cost. Algorithmic on-chip access control list (ACL) capability is available along with seamless expansion using Broadcom's industry-leading NL566xx knowledge based processor (KBP).

The Network of the Future - Delivered Today
Broadcom's NPU solutions are enabling the service provider network of the future today, providing the enhanced high-bandwidth capacity and scale to meet the aggressive requirements of next generation networks driven by the insatiable demand for data. With a fully programmable forwarding architecture, the BCM88030 family allows network operators to reduce CAPEX by adapting to emerging protocols without the costly redeployment of network equipment.

Spanning the entire network - from the access point to the edge, to the aggregation and finally the core - Broadcom high bandwidth solutions deliver the mobile experience consumers demand - with high quality voice connections, faster application downloads, and uninterrupted video streaming.

Market Drivers:

• Global IP traffic expected to grow 18-fold by 2016(1)
• Carrier Ethernet equipment market to reach $40.2 billion by 2015(2)
• Massive traffic growth challenges current network architecture(3)
• Explosive bandwidth growth and the rapid adoption of 40GbE and 100GbE in service provider networks(3)
• Emerging and changing protocols driving the need for programmability in the data path

Key Features:

• 64 custom processors operating at 1 GHz provide 100GbE carrier switching and routing with full programmability
• Advanced software development environment and feature-rich carrier Ethernet switch/router application enables robust user programmability and fast time-to-market
• Proprietary look-up algorithm utilizes low-cost DDR3 DRAM for highly scalable MAC & IP tables while dramatically lowering overall system cost
• Integrated OA&M engine, 1588 timing and synchronous Ethernet, 10GbE SerDes and 1/10/40/100GbE Ethernet MACs substantially reduces board space, system cost and power
• Seamless integration with XLP multi-core processors for best-in-class control plane processing

Availability
The BCM88030 family consists of three devices, including the 100 Gbps BCM88038 NPU, the 50 Gbps BCM88034 NPU and the BCM88032 24 Gbps NPU. All devices are now sampling with production volume slated for the second half of 2012.(1) Cisco Visual Networking Index: Forecast and Methodology, 2011-2016
(2) Infonetics Service Provider Router & Switch Forecast, 4Q11
(3) Linley Guide to Network Processors, April 2011