| Order a report

A Guide to Network Processors

Sixteenth Edition

Published November 2015

Authors: Bob Wheeler and Loring Wirbel

Single License: $4,495 (single copy, one user)
Corporate License: $5,995

Ordering Information



The Definitive Report on Network Processing

NPUs are a critical ingredient of carrier-equipment designs. These chips appear in designs from leading OEMs spanning many applications from wireless backhaul to Carrier-Ethernet switch/routers. This broad adoption created a merchant market that exceeded $300 million in 2014, large enough to sustain multiple vendors. In 2015, however, three of the active NPU vendors are being acquired, service providers are adopting NFV, and NPUs are finding new applications in data centers.

High-end NPUs are scaling to 200Gbps and beyond to support high-density line cards and 100G Ethernet. Some remain focused on Layer 2/3 forwarding, whereas others are addressing flow processing at Layers 4 through 7. The latter group is also addressing new system architectures including smart top-of-rack switches and appliances using intelligent NICs (iNICs). A "Guide to Network Processors" provides a single comprehensive report covering NPUs spanning carrier and data-center applications.

This report covers programmable NPUs including: Broadcom’s XGS Core and StrataDNX lines, EZchip’s NP and NPS lines, Netronome's NFP, and PMC-Sierra’s WinPath. New to this edition is coverage of iNICs from EZchip (Tilera) and Netronome. This report also includes coverage of search coprocessors, such as Broadcom’s KBP products, Cavium’s Neuron Search, Marvell’s Questflo, and Renesas’s TCAMs.

Only The Linley Group follows this market closely enough to give you the complete picture. Which vendors are in this business for the long haul? How do the latest products stack up? What is the market outlook for merchant NPUs? "A Guide to Network Processors" is the result more than a decade of research that cannot be duplicated. If you are interested in following this strategic standard-product segment, you have located the definitive source.

Get Facts, Not Fiction

This report cuts through the vendor hype and gives you the solid information you need to understand this market. "A Guide to Network Processors" analyzes each vendor and each product, probes their strengths and weaknesses, then presents key details in a consistent, easy to compare fashion. For those less familiar with this combination of networking and CPU design, the report includes several introductory chapters that define and describe basic concepts and key technologies.

The Linley Group is the most recognized and respected name for technology and market analysis of network processors. Don't be fooled by weak overviews written by market analysts who really don't understand how a network processor works. The "Guide" provides a unique combination of business and technology savvy from the leading analysts in this market. Bob Wheeler and Loring Wirbel use their long experience in the networking world to analyze these devices. Together, the two authors ignore the fiction and provide the real story on each NPU vendor and its products.

Don't miss the latest information on this important market. Order now.

This report is written for:

  • Engineers who are designing carrier-networking equipment and need to select a network processor (NPU)
  • Network architects at carriers and service providers who need to get up to speed on this technology
  • Marketing and engineering staff at companies that sell NPU products or partner with NPU vendors
  • Technology professionals who wish an introduction to carrier-networking and network processors
  • Financial analysts who desire a detailed analysis and comparison of NPU companies and their chances of success
  • Universities and research institutes that need a vendor-neutral introduction to this market and technology

What's New in This Edition

Updates to the 16th Edition of "A Guide to Network Processors"

"A Guide to Network Processors" has been extensively updated to include the latest disclosures from NPU vendors as well as 2015 market data.

Here are some of the many changes you will find:

  • New quantitative market data, including:
    • Preliminary NPU vendor market shares for 2015
    • Preliminary search coprocessor vendor market shares for 2015
    • Revised forecasts for merchant NPUs and search coprocessors through 2020
  • Coverage of Broadcom’s 28nm StrataDNX (Dune) line
  • New coverage of EZchip’s TILEncore intelligent NICs
  • New coverage of Netronome’s FlowNICs
  • Extensive updates to company-background information, roadmaps, and analysis for major NPU vendors
  • Revised and updated tutorials

The viability of the merchant NPU market is the result of adoption by Tier One OEMs across a broad range of network-equipment segments. In 2015, we believe all Tier One OEMs are using merchant network processors in selected designs. Even those with internally developed NPUs — such as Alcatel-Lucent, Cisco, Ericsson, Huawei, and Juniper — are using merchant products in some designs. These OEMs, however, are continuing to pursue internal development, limiting the growth rate for high-end merchant NPUs.

Although the network-processor market as a whole suffered a four-year revenue decline in 2011–2014, the top-line view does not tell the whole story. Revenue from high-end NPUs has been growing since 2012, whereas sales of access NPUs have been in decline. Revenue from high-end NPUs grew 10% in 2014, while revenue for access NPUs fell 10%. Shipments of high-end NPUs have been growing thanks to demand for service-provider routers. By contrast, access NPUs declined as the transition from 3G to 4G (LTE) wireless technology reduced demand from base stations and RNCs. Looking forward, we expect the NPU market will return to significant growth in 2016 as the access segment becomes a smaller contributor and high-end shipments continue to rise.

EZchip is the leading supplier of high-end merchant NPUs, and larger Mellanox is acquiring it. The company is shipping multiple product generations including the 200Gbps NP-5, which combines 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 400Gbps NPU, called the NPS-400, that handles Layer 4–7 features and offers easier programming than traditional NPUs. It’s also shipping intelligent NICs based on processors from Tilera, which EZchip acquired in November 2014.

Broadcom is shipping NPUs in two product lines, which have their origins in two different acquisitions. From the Sandburst heritage, the company is shipping the customer-programmable BCM88038 — the industry’s first full-duplex 100Gbps NPU. In the former Dune line (now branded StrataDNX) Broadcom is sampling the BCM88670, which delivers an industry-leading 720Gbps of throughput. Although the BCM88670 is internally programmable, the company sells it as a configurable device using factory-supplied microcode. The StrataDNX family includes a complete switch fabric that works with both Broadcom NPUs. The company is also the leading vendor of search coprocessors with its KBP family. Thanks to its broad offering, Broadcom provides system solutions that are more complete than any competitor’s.

Netronome is now the only privately held company developing NPU silicon, but it primarily delivers these chips as part of intelligent NICs. Rather than competing for carrier-equipment designs, it principally targets security appliances and other data-center applications. The company shipped its 40Gbps NFP-3240 and associated iNICs in 2011. In 2015, Netronome shipped the NFP-6xxx and new FlowNICs that scale throughput to 200Gbps. Designed for stateful flow processing, the NFP devices were the industry’s first NPUs to perform advanced services such as IPSec, SSL, firewall/NAT, load balancing, and deep packet inspection.

In the access market, PMC-Sierra is the leading vendor. The company entered the NPU market in 2010 by acquiring Wintegra, a leading vendor of access NPUs because of its flexible architecture and complete data-plane software. Following the acquisition, PMC focused on wireless-backhaul designs, where support for legacy protocols continues to differentiate the WinPath NPUs from Ethernet-only alternatives. In 2014, PMC began production of WinPath4, which scales performance to 40Gbps.

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 supplier of search coprocessors. Its traditional competitor is Renesas, which primarily supplies TCAMs to Cisco. Two new entrants, Cavium and Marvell, hope to compete with Broadcom using algorithmic implementations that promise to reduce power. Depending on their application, OEMs now have at least two vendors to choose from.

List of Figures
List of Tables
About the Authors
About the Publisher
Preface
Executive Summary
1 Introduction to Carrier Networks
Network Types and Topologies
Metro-Area Networks (MANs)
Wireline Access
Wireless Access
SDN and NFV
Equipment Types
Metro Platforms
Wireline Access Infrastructure
Wireless Access Intfrastructure
2 Carrier-Network Technology
Network Layers and the OSI Model
Layers 3–7
Layers 1 and 2
Network Protocol Details
Ethernet
PPP
Multiprotocol Label Switching (MPLS)
Sonet/SDH
OTN
IP Multicasting
TDM Emulation
Timing Synchronization
Ethernet OAM and Protection Switching
HDLC and Frame Relay
Carrier Ethernet Services
Software-Defined Networks
Network Functions Virtualization
OpenFlow
Open vSwitch
Tunneling Protocols
Carrier Ethernet Services
Packet-Processing Pipeline
Control and Data Planes
Parsing
Classification
Forwarding
Modification
Network Paths and Quality of Service
Traffic Management
Policing and Shaping
Congestion Management
Scheduling
Hierarchical Traffic Management
Network and I/O Interfaces
Ethernet PHY Interfaces
Interlaken and Interlaken Look-Aside
OIF/NPF LA-1
Utopia, POS-PHY, and SPI
PCI Express
Chassis and Board Standards
3 NPUs, Search Coprocessors, and iNICs
What Is a Network Processor?
What Is Not a Network Processor?
Common NPU Characteristics
Microarchitecture Variations
Encryption Engines
Fixed-Function Versus Programmable
Network Interfaces
Memory Interfaces
Host Interface
Software Considerations
Vendor Programming Versus Customer Programming
Search Coprocessors
What Is a Search Coprocessor?
Applications
Architecture and Common Characteristics
Intelligent NICs
4 Market Size and Trends
Merchant-NPU Market Size and Segmentation
NPU Market Share by Vendor
NPU Market Forecast
Search-Coprocessor Market
Technology Trends
OEM-Proprietary NPUs
Alcatel-Lucent FlexPath
Cisco QuantumFlow Processor and nPower X1
Ericsson SNP-4000
Huawei Solar
Juniper Junos Trio
Network Processing Technology Trends
SDN and NFV
Stateful Network Processors
Multicore Processors Versus NPUs
FPGAs Versus NPUs
Intelligent NICs Versus Line Cards
5 Broadcom
Company Background
Key Features and Performance
Packet Processors
Search Coprocessors
Design Details
Packet Processors
Search Coprocessors
System Design
Development Tools
Product Roadmap
Conclusions
6 EZchip
Company Background
Network Processors
Key Features and Performance
Internal Architecture
System Design
Intelligent NICs
Development Tools
Product Roadmap
Conclusions
7 Netronome
Company Background
Flow Processors
Intelligent NICs
Software Drivers and Tools
Product Roadmap
Conclusions
8 PMC-Sierra
Company Background
Key Features and Performance
Internal Architecture
System Design
Development Tools
Conclusions
9 Search-Coprocessor Vendors
Cavium
Company Background
Key Features and Performance
Conclusions
Marvell
Company Background
Search Coprocessors
Conclusions
Renesas
XeL Technology
10 NPU and iNIC Comparisons
200Gbps-and-Above NPUs
40Gbps-and-Above iNICs
11 Conclusions
NPU-Vendor Outlook
EZchip
Broadcom
Netronome
PMC-Sierra
Closing Thoughts
Appendix: Further Reading
Index
Figure 1‑1. Generic network architecture.
Figure 1‑2. Wireline access networks.
Figure 1‑3. 3GPP Release 8 LTE network architecture.
Figure 2‑1. MPLS encapsulation.
Figure 2‑2. VPLS switch conceptual model.
Figure 2‑3. Control and data planes.
Figure 2‑4. Hierarchical traffic-management example.
Figure 2‑5. Standard line-card interfaces.
Figure 3‑1. Block diagram of a typical NPU.
Figure 3‑2. Hybrid search-engine architecture.
Figure 4‑1. NPU revenue by application segment, 2015.
Figure 4‑2. Preliminary market share of merchant NPUs by revenue, 2015.
Figure 4‑3. Merchant NPU market forecast, 2014–2020.
Figure 4‑4. Merchant search-coprocessor market forecast, 2014–2020.
Figure 5‑1. Block diagram of Broadcom BCM8803x NPU.
Figure 5‑2. Block diagram of Broadcom BCM88670.
Figure 5-3. Block diagram of Broadcom NLA12k KBP.
Figure 5‑4. Block diagram of a Broadcom-based 400Gbps line card.
Figure 6‑1. Block diagram of NPS-based network-processing cluster.
Figure 6‑2. Block diagram of EZchip NP‑5.
Figure 6‑3. Programmable top-of-rack switch using EZchip NPS-400.
Figure 7‑1. Block diagram of Netronome FlowNIC-2x100.
Figure 8‑1. Block diagram of PMC-Sierra WinPath4.
Figure 8‑2. Wireless-backhaul aggregator based on WinPath4 and UFE4.
Figure 8‑3. Microwave backhaul using WinPath3.
Table 2‑1. OSI reference model.
Table 2‑2. Bandwidths of common interfaces.
Table 4‑1. Worldwide revenue of merchant NPUs by vendor, 2014–2015.
Table 4‑2. Worldwide revenue of merchant search coprocessors by vendor, 2014–2015.
Table 5‑1. Key parameters for Broadcom NPU and TM devices.
Table 5‑2. Key parameters for Broadcom KBP devices.
Table 6‑1. Key parameters for EZchip NPUs.
Table 7‑1. Key parameters for Netronome NFP-3240 and NFP-6xxx.
Table 8‑1. Key parameters for PMC-Sierra WinPath devices.
Table 8‑2. WinPath protocol and interworking support.
Table 9‑1. Key parameters for XeL ISE76K search coprocessor.
Table 10‑1. Comparison of 200Gbps-and-above NPUs.
Table 10‑2. Comparison of 40Gbps-and-above intelligent NICs.

Events

Linley Processor Conference 2016
Focusing on embedded processors for communications, automotive, and IoT
September 27 - 28, 2016
Hyatt Regency, Santa Clara, CA
Register Now!
More Events »

Newsletters

Linley Wire
Analysis and news on processors for networking and communications
Linley on Mobile
Analysis and news on semiconductors for mobile and wireless
Processor Watch
Analysis of high-performance microprocessor developments
Subscribe to our Newsletters »