interop 2006: evolution of the networking industry

Post on 02-Dec-2014

157 Views

Category:

Technology

1 Downloads

Preview:

Click to see full reader

DESCRIPTION

This was a presentation from 2006 where (starting on slide 30) I described how the way consumer networking vendors built their products would ultimately upend the way enterprise vendors built products. Today, the hype around SDN, commodity switching products, and linux based network operating systems is making much of this come true. This presentation was first given in May 2006 at the Interop trade show. Slide 30, is where I talked about how the networking market would split from vertically organized vendors to horizontally specialized vendors. I predicted this transition would take 10 years. At the time, the use of merchant silicon in datacenter and enterprise switching products was a rarity. Today, it is common. Then linux derivative network OSes were limited to consumer products, today they are taking over large scale cloud datacenters. The networking market hasn’t come as far as I expected in 10 years, but it’s pretty cool much of this either came true or is on the cusp of coming true. Much more change is under way.

TRANSCRIPT

www.idc.com

The State of Enterprise Networking: Interop 2006

Abner GermanowDirector, Enterprise Networking

Agenda

• A quick refresher on disruptive markets• The state of the enterprise networking market• Solving the complexity and crisis• Guidance

©IDC, 2006

A quick refresher course

in disruptive markets

Sustaining vs. Disruptive Innovation

Performance that customers

can utilize or absorb

New performance trajectory

Disruptive Innovation

Pace of

technological

progress

Sustaining Innovation

Breakthrough

Incremental

Pro

du

ct

Pe

rfo

rma

nc

e

TimeSource: The Innovator’s Dilemma

Interdependency Vs. Modularity

Modular Archite

ctures

Beat competitors with speed, responsiveness and standards

Pe

rfo

rma

nc

e

Time

Interdependent Archite

ctures

Beat competitorswith complete solutions

Shifts in Interdependency and Modularity Has a Long IT History

Equipment

Materials

Components

Product design

Operating system

Applications software

Sales & distribution

Field service

Intel, Micron, Quantum, Komag, etc.

Compaq, Dell, Gateway, Packard Bell

Microsoft

Word Perfect, Lotus, Borland

CompUSA

Independent contractors

Microsoft

1960–1980 1980–1990 1990–Present

Dell

IBM

Co

ntr

ol D

ata

Dig

ital E

qui

pm

ent

Monsanto, Sumitomo Metals, Shipley, etc.

Teradyne, Nikon, Canon, Applied Materials, Millipore, etc.

Assembly Compaq Contract Assemblers

Where are we today?

Gigabit is healthy, but 100 MB will represent the majority of shipments for a while

0

2,000,000

4,000,000

6,000,000

8,000,000

10,000,000

12,000,000

14,000,000

16,000,000

1Q02 2Q02 3Q02 4Q02 1Q03 2Q03 3Q03 4Q03 1Q04 2Q04 3Q04 4Q04 1Q05 2Q05 3Q05 4Q05

2002 2003 2004 2005

fix m - 100 MB

fix m - 1000 MB

mod - 100 MB

mod - 1000 MB

Layer (All) Vendor (All)

Sum of Ports

Year Quarter

Form Factor

Speed

10 G Port Growth

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

1Q02 2Q02 3Q02 4Q02 1Q03 2Q03 3Q03 4Q03 1Q04 2Q04 3Q04 4Q04 1Q05 2Q05 3Q05 4Q05

2002 2003 2004 2005

fix m

mod

Layer (All) Vendor (All) Speed 10 G

Sum of Ports

Year Quarter

Form Factor

WLAN Market Share by Shipments

0%

20%

40%

60%

80%

100%

4Q2004

1Q2005

2Q2005

3Q2005

4Q2005

Other

3Com

Aruba

Symbol

Cisco

WLAN Share of Shipments

0%

20%

40%

60%

80%

100%

4Q2004

1Q2005

2Q2005

3Q2005

4Q2005

AP vs Controller Share by Revenue

Controller

AP

WLAN Product Mix Shifted: Share of Revenue

Where are we going?

Core Enterprise Network Forecast

$0

$2

$4

$6

$8

$10

$12

$14

$16

$18

2003 2004 2005 2006 2007 2008 2009

Billions

LAN Switch Router

New Enterprise Markets Are Hot

$0$5

$10$15$20$25$30$35$40

2003 2004 2005 2006 2007 2008 2009 2010

Billions

LAN switch Router IP PBX WLAN Application Networking

What’s on the agenda in 2006?

0% 5% 10% 15% 20% 25% 30%

Branch Routers

Compliance

Wiring closet LAN switch

Network Management

Core/Backbone LAN switch

Headquarters Routers

IP Voice

No major networking initiatives

Wireless LAN deployments

Security function / appliances

Source: IDC Enterprise Networks Vertical Views Survey, May 2006 N=623 (Prelim data, unweighted)

Q: What are the top two upgrades or initiatives making up largest share of your organization's data network equipment spending over the next 12 months?

Security

Problems: 1. Need to reduce response times, but security and

networking groups often have different priorities2. Conga line of appliances doesn’t scale

©IDC, 2006

Security goal #1

Reduce response times

Network

Network identifies suspect traffic

Network

Network identifies suspect traffic

Security Function

Security productanalyzes traffic

Security Function

Network

Security productanalyzes traffic

Security productissues mitigationrecommendations

to network

Network identifies suspect traffic

Network

Network identifies suspect traffic

Network takes action to block,

rate limit, or quarantine

Security Function

Security productanalyzes traffic

Security productissues mitigationrecommendations

to network

Demand For Closed Loops Exists In All Hot Markets

SecurityMobility

IP Telephony

App Networking

Enterprise Network

StorageNetwork

Compute

Storage

Common Themes In Hot Topics

A closed loop with the network is a win-win

Control may lie outside networking group

Web services standards

SecurityMobility

IP Telephony

App Networking

Enterprise Network

StorageNetwork

Compute

Storage

One End of the Solution Spectrum: Build It Yourself

Acquire or develop technologies and integrate in house

Pros: Grow revenues through growth in hot products. Development teams can be forced to work together

Cons: Assumes customer wants to buy everything from you Angry partners

©IDC, 2006

The Other End of the Spectrum: Play with Others

Develop a set of APIs and use a standard platform build links to secondary and tertiary technologies

Pros: Customers can use multiple vendors Vendors can specialize

Cons: Customers can’t exploit the value of the open platform until

the 2nd or 3rd application Multiple vendors blame others for problems

©IDC, 2006

Security goal #2

Consolidate conga line of appliances

Classes of Appliances at the WAN gateway

• WLAN access point management• Wide area file services• WAN Bandwidth Management• URL Filtering/content management• IP PBX/PSTN Gateway• Network access control/computer quarantine• Intrusion detection/prevention• Messaging Security• Authorization/authentication• Anti-virus/spyware

What platforms do customers want a VPN to run on?

16%

43%6%

25%

10%General purposeserver

Purpose builtappliance

Blade in blade server

LAN Switch or router

Managed Service

Q14. I am going to read a list of applications, and I would like you to indicate your preference of five platforms for that specific application. You can select more than one platform for each application. In other words, on which platform would you like … to reside?

Source: IDC Next Generation Network and Security Special Study, May 2006 N=411

COTs Modularity:Breaking Networking into

IT-sized Chunks

Shifts in Interdependency and Modularity Has a Long IT History

Equipment

Materials

Components

Product design

Operating system

Applications software

Sales & distribution

Field service

Intel, Micron, Quantum, Komag, etc.

Compaq, Dell, Gateway, Packard Bell

Microsoft

Word Perfect, Lotus, Borland

CompUSA

Independent contractors

Microsoft

1960–1980 1980–1990 1990–Present

Dell

IBM

Co

ntr

ol D

ata

Dig

ital E

qui

pm

ent

Monsanto, Sumitomo Metals, Shipley, etc.

Teradyne, Nikon, Canon, Applied Materials, Millipore, etc.

Assembly Compaq Contract Assemblers

What will networking look like?

Equipment

Materials

Components

Hardware design

Operating system

Applications software

Sales & distribution

Field service

1990–Present

Assembly

Usual Suspects

Usual Suspects

The Future

3C

om

, Cis

co, N

ort

elAtheros, Broadcom, Intel, Marvel, ??

Contract Assemblers

Accenture, IBM, EDS, VARs, SIs

Accenture, IBM, EDS, VARs, SIs

Atheros, Broadcom, Intel, Marvel, ??

What will networking look like?

Equipment

Materials

Components

Hardware design

Operating system

Applications software

Sales & distribution

Field service

1990–Present

Assembly3

Co

m, C

isco

, No

rtel

Intoto? Nexthop? Level 7? Symbol?

Linux?, IOS? JUNOS? Others?

Cisco? Extreme? IBM BladeCenter? Bivio?

Contract Assemblers

Usual Suspects

Usual Suspects

3Com? Adtran? Cisco? Procurve?

The Future

Example: Intoto Ecosystem

Source: Intoto

OS Vendors(BSP integration)

SnapGear

3rd-Party Apps

Antivirus

Antispam

Content filtering

WLAN autoconfig.

Hardware ODM

Semiconductor Vendors (Chip integration)

Service Providers (Provisioning and certifications)

3rd-Party Certification Consortiums(Industry Standard Certifications)

End-user Product(OEM Branding + Channel + Support)Networking OEM

Hardware ODMHardware Platform

(CPU/NP, coprocessors, PCBA, OS, and BSP)

Production-ready Security Gateway Platform

(Intoto Network-centric Security Gateway Platform Software + Integration + Certifications)

Software ODM

Modularity in Networking: New Market Disruption

Example: Netgear

Buy code from third parties

Off-the-shelf silicon

Off-the-shelf operating systems

Integrate software and hardware

ODM manufacturing

Consumer & SMB Focus

Low-End Disruption

Target over-served customers

with lower-cost business model

Performance

Time

Sustaining Innovation

Bring better products into

established markets

Three Strategies for Market Disruption

Non-consumers orNon-consuming

occasionsTime

Different Performance

Measure

Low-End Disruption

Target over-served customers

with lower-cost business model

Performance

Time

Sustaining Innovation

Bring better products into

established markets

New Market Disruption

Compete against

non-consumption

Three Strategies for Creating Growth

New Market Space: Consumer & Security Products

Long-term Outlook for Network COTs Modularity

Traditionalnetwork

R&D

OSASICs

Long-term Outlook for Network COTs Modularity

COTS model matures in consumer and SMB markets

TraditionalNEM R&D

ModularNEM R&D

OSASICs

IntegrationUsabilityPartners

Specialists

3 Notes on Market Disruptions

Disruptions in networking take a long time (10+ Years) Disruptive businesses are often lower margin than

incumbents COTS model comes with risk that incumbent and new

entrants fail

Networking Vendors:Where Are They Headed?

Start-ups

Airspace, Andiamo (both acquired by Cisco,) Aruba, NetDevices, Trapeze, many others

Develop on Linux

Off-the-shelf silicon

Consolidate Apps Functions Services

©IDC, 2006

Disruptors

Adtran, Dell, D-Link, Huawei?, F5, Linksys (Acquired by Cisco), Netgear, ZTE?

Off the shelf silicon (Broadcom, Intel, Marvel, etc)

Network device software Intoto, Level 7, Nexthop

Customer-defined controls (F5)

OS: Windriver, Linux, other COTS OS

©IDC, 2006

Challengers

3Com, Enterasys, Extreme, Foundry, Juniper, Nortel, Procurve

Mix of custom ASICs and COTS silicon

OEM or partner with specialists for security, voice, WLAN

Open APIs to internal developers and/or partners

Modular operating systems

©IDC, 2006

Modularity in Networking: Cisco’s Opportunity & Threat

Integration is a key value proposition for voice, security, and mobility

Little to no incentive to buy COTs

IOS is now “internally modular”

Will Cisco’s ability to add, consolidate, and simplify new features or even acquire and integrate become too slow?

Cisco

IIN: Intelligent Information Networking

Technologies that produce a better network

SONA: Services Oriented Network Architecture

An architecture and technology set that should enable customers to virtualize and consolidate a set of infrastructure services across an enterprise.

The benefit of this consistency will be to helping other parts of IT gain efficiencies and increase the network’s share of total IT spend.

Guidance For Vendors

1. The enterprise network market continues to grow, with new access, WAN, and applications driving demand for new functions, scale, and manageability

2. COTs does not equal commodization of network equipment in the enterprise.

3. Enterprise equipment vendors need to use COTs in order to shift and prioritize R&D, not to cut R&D spend

4. There is resistance to network service integration INTO network equipment, but demand for integration with in the near term WITH network equipment

Contact Info

Please email me atagermanow@idc.com

top related