All Tier One networking and communications OEMs use NPUs in a range of systems, including edge/services routers, Carrier Ethernet switch/ routers (CESRs), optical-transport platforms, and wire¬less and broad¬band infrastructure. Many of these OEMs continue to invest in internal NPUs, which limit the adoption of merchant products. But most of those with internally developed NPUs still use merchant NPUs in some designs.
Because semiconductors are subject to occasional inventory corrections, the merchant-NPU market has dis¬played a somewhat volatile growth rate over the past decade. This trend continued through the recession of 2009, followed by an upswing in 2010. What followed, however, was an un¬pre-cedented two-year decline in 2011–2012. This decline was due to several factors including inventory correc¬tions, weak macroeconomic conditions, and declining revenue from ancillary markets. But the merchant-NPU market was still about $100 million larger in 2012 than it was in 2007, and we predict it will reach nearly half a billion dollars in 2017.
Despite being smaller than most competitors, EZchip is the leading sup-plier of high-end merchant NPUs. The company is shipping mul¬tiple product generations including the 100Gbps NP-4. EZchip is also sampling the NP-5, the industry's first 200Gbps chip combining a customer-programmable packet processor, traffic manager, and Ethernet MACs. It has worked with Marvell to supply special versions of its NPUs to Cisco, which primarily uses them in the ASR 9000 router. EZchip is developing a new line of NPUs, called NPS, that handle Layer 4–7 features and promise easier programming than traditional NPUs.
In the access market, PMC-Sierra is the leading vendor. The company entered the NPU market in late 2010 by acquiring Wintegra, which had become a leading vendor of access NPUs because of its flexible archi-tecture and complete data-plane software. Since the acquisition, PMC has increasingly focused on wireless-backhaul designs, where support for legacy protocols continues to differentiate the WinPath NPUs from Ethernet/IP-only NPUs. Secondary markets for WinPath include wireless base stations and broadband-aggregation equipment. In 2013, PMC is sampling WinPath4, which scales performance to 40Gbps.
With its BCM88030, Broadcom is shipping the industry's first full-duplex 100Gbps NPU. The company combines this NPU with its 200Gbps configurable traffic manager, switch fabric, and search coprocessors to offer a packet-processing solution that is more complete than any com-petitor's. Customers can combine these components in 400Gbps line cards or use them independently in different designs. Broadcom also provides ASIC services to OEMs wishing to develop in-house NPUs, such as Alcatel-Lucent with its FP3. For access and aggregation applications, it instead offers configurable Carrier Ethernet switch chips in its popular StrataXGS line.
In addition to its manufacturing role with EZchip, Marvell supplies high-end NPUs of its own design. These products come from the company's acquisition of NPU startup Xelerated. The company's flagship NPU, the HX, can support 100Gbps line cards as well as standalone pizza-box configura¬tions. The derivative AX line delivers programmable packet processing for access- and aggregation-switch designs. Marvell has also added traffic-management technology from its NPUs to configurable Carrier Ethernet switch chips, which address lower-cost access designs.
LSI is an incumbent vendor of access NPUs but has taken a new direction with its Axxia com¬mu¬nications processors. Combining mul¬tiple general-purpose CPUs with data-plane technology from the company's earlier APP line, the Axxia devices represent a hybrid of a multicore processor and an NPU. Rather than targeting NPU applications, however, Axxia pri¬marily targets wireless base stations. Through its ASIC operations, LSI offers semicustom devices that can integrate baseband pro¬cessing.
Netronome is now the only privately held company developing NPU sili-con. It started to generate NPU revenue in 2011 with initial shipments of its 40Gbps NFP-3240. Although this device cannot match the throughput of high-end NPUs, it is unique in its ability to perform advanced services such as IPSec, SSL, firewall/NAT, load balancing, and deep packet inspection (DPI). Rather than competing for carrier designs, Netronome principally targets security appliances and other data-center applications. In 2014, Netronome plans to sample the NFP-6xxx, its next-generation NPU that will scale throughput to 200Gbps.
Also referred to as TCAMs or network search engines (NSEs), search coprocessors offload NPUs by performing lookup functions. Cisco and other leading OEMs combine search coprocessors with both merchant and internal NPUs, resulting in a merchant search-coprocessor market about two-thirds the size of the merchant-NPU market. As with NPUs, some OEMs also possess proprietary search designs.
Through its NetLogic acquisition, Broadcom became the dominant sup-plier of search coprocessors, which the company calls knowledge-based processors (KBPs). Its traditional competitor, Renesas, has fallen behind in technol¬ogy and is losing market share. Two new entrants, Cavium and XeL Technology, hope to compete with Broadcom using algorithmic implementations that promise to reduce power and cost. By 2015, OEMs should have several competitive vendors to choose from.