Semiconductor Industry - Needs more than Moore's Law

IBM Global Business Services

IBM Institute for Business Value

Electronics
More than
“Moore” to win
Optimization strategies
for success in a maturing
semiconductor industry

IBM Institute for Business Value
IBM Global Business Services, through the IBM Institute for Business Value,
develops fact-based strategic insights for senior executives around critical public
and private sector issues. This executive brief is based on an in-depth study by
the Institute’s research team. It is part of an ongoing commitment by IBM Global
Business Services to provide analysis and viewpoints that help companies realize
business value. You may contact the authors or send an e-mail to iibv@us.ibm.com
for more information.

More than "Moore" to win
Optimization strategies for success in a maturing semiconductor industry
By George Bailey and Wendy Huang

The impact of higher costs and other key disruptive trends on the semiconductor
industry – especially competitive landscape changes, technology convergence
and greater global connectedness – mean that traditional business models just no
longer work. Nor will the promises of Moore’s Law be enough to provide sustained
competitiveness. Future success will require innovative changes to existing
business models that optimize capabilities in the areas of integration, customer
centricity and collaboration.
Industry costs are skyrocketing. R&D costs and the five-year compound annual growth
are expected to rise by a compound annual rate (CAGR) for industry revenue dropped
rate of 12.2 percent per year between 2004 from more than 15 percent in 1990 to less than
and 2010, while the industry revenue growth 4 percent by 2005.
1
rate remains only 6.0 percent per year. The
total R&D cost for the semiconductor industry FIGURE 1.
Five-year CAGR of semiconductor industry revenue,
was reported to be US$45 billion in 2006 1985 to 2005.
and analysts believe that it will reach US$100
2 US$ billion Percent
billion by 2010. In addition, process develop- 250,000 20
ment costs for 32nm manufacturing could hit 200,000 15
US$3 billion, which is twice the cost for 65nm 150,000
3
10
process technologies. This does not include 100,000
50,000 5
other development costs such as new chip
0 0
fabrication facilities (“fabs”), or new processes, 1985 1990 1995 2000 2005
tools and equipments. Sales
5 year CAGR
Market indicators show the industry is, in fact, Source: IBM Institute for Business Value analysis of Semiconductor
Industry Association/SICAS data. https://www.sia-online.org/down-
maturing. Revenue growth has slowed dramat-
loads/shares.pdf.
ically in the past two decades (see Figure 1),

1 More than “Moore” to win

And yet, in today’s world of pervasive To better understand current industry trends
computing, it seems semiconductor devices, and their impacts, as well as how to win in this
or “chips,” can be embedded practically new environment, IBM conducted the 2007
everywhere: in vehicles to monitor the need for Semiconductor Industry Optimization Study
replacement parts, in humans to detect heart- (SOS). SOS looked in detail at the top 60 semi-
beats and brain activity, and even in growing conductor suppliers by size of revenue, which
plants to communicate needs for water or includes a mix of companies. Some of the top
fertilizer. With no end in sight to “ubiquitous 60 design and manufacture their own chips
computing” and “embedded networks and (Integrated Device Makers or “IDMs”), some
controls,” these examples are but the tip of the design, but do not manufacture chips (fabless)
iceberg of potentially endless opportunities for and four top companies manufacture, but do
the semiconductor industry. not design chips (foundry players). Our study
included face-to-face interviews, analysis of
Historically, Moore’s Law offered strong hope financial results and industry best practices,
of continual cost reduction and continued along with substantial secondary research.
prosperity for chip makers in the face of ever-
increasing demand. The1965 prediction by SOS results led to some stark conclusions for
Gordon Moore of Intel postulated that the this maturing industry. No longer can competi-
number of transistors on a chip would double tive advantage be found solely in continued
every one to two years due to advances in technology advancement, as espoused by
4
technology. With today’s explosive use of Moore’s Law. The growing pervasiveness of
semiconductor technology in all walks of life, chip use across industry and society requires
the current reality is that it now takes more companies to re-examine their respective
than “Moore” to win. business strategies and supporting business
models to properly adjust for the dynamic
Today, rumors of mergers and acquisitions competitive landscape, increasing technology
(M&As) run rampant in the chip industry. In convergence and growing global connected-
recent years, the increased costs of staying ness. Only well-considered choices can lead
competitive have driven many companies to the stable financial results necessary for
to increase their collaboration with peers, in survival and success as the industry faces
forms ranging from joint development to M&A. significant disruption.
Companies failing to properly manage their
bottom-line financials are increasingly targets
for takeover, especially as private equity firms
take a more active role in the industry.

2
2 IBM Global Business Services
IBM Global Business Services

More than “Moore” to win
Optimization strategies for success in a maturing semiconductor industry

The seven Cs: Today’s disruptive developments require at least two mask sets
industry trends to get to production, this can put a big dent in
Along with industry maturation, seven disrup- the semiconductor suppliers’ wallets.
tive trends – we call them the “seven Cs”
Competition: An industry in flux thanks to
– are reshaping industry players, as well as
China, India and private equity
their relationships within the ecosystem and
The competitive landscape is constantly
shifting the value chain itself. And like the
changing. Within ten years, it is estimated that
“seven seas” of Earth, these seven Cs may
40 percent of today’s semiconductor vendors
require special attention to navigate choppy
are likely to leave the industry as a result of
waters ahead. These disruptive semiconductor 9
intense global competition. The number of
industry forces include complexity, commod-
Chinese fabless companies increased from 15
itization and consumerization, along with four
in 1990, to an astounding 479 in 2005 and still
others that this paper will examine much more 10
growing.
closely: cost, competition, convergence and
connectedness. India, too, is a rising contender. India boasts
125 fabless companies that have grown from
Cost: Doing business is more expensive
around US$1 billion to US$3 billion, and they
across the board
are slated to reach US$40 billion in sales
In virtually every dimension of the semi- 11
by 2015. Most of these Chinese and Indian
conductor industry, costs are increasing
companies are not well known, but it is only a
exponentially. With each successive leading
matter of time before some will become cred-
technology, the level of complexity grows in
ible contenders.
conjunction with the cost. R&D costs asso-
ciated with leading technology IC design, Private equity activities have also intensified
process and equipment can be attributed to in the past few years. The high-profile lever-
the escalating cost of semiconductor industry aged buyouts of NXP and Freescale alone
R&D. According to analysts, process tech- accounted for US$27 billion in 2006 (See
nology R&D costs alone jumps from US$2.4 Figure 2). The private equity playbook consists
billion at the 45-nm node to US$3 billion at of first leveraging the strong cash position
5
32 nm. A 300-mm, 45-nm wafer is about 10 of the target company to borrow money.
times more expensive than a 200-mm, 250-nm Next, investors restructure and improve the
6
wafer. company’s bottom line by driving inefficiency
out of the business, then later either sell the
Mask costs also jumps significantly with each
company or take it public. Such deals have
leading edge technology. For instance, mask
been able to generate returns of 30 percent to
costs at 90-nm are about US$800,000 – at the 12
7 40 percent. As a result, private equity-owned
65-nm node, they reach about US$1.2 million.
companies are setting new business operation
For a 45-nm node, mask costs alone can total
8 standards for all of the industry players.
more than US$2 million. Because most chip


Traditional FIGURE 2.
semiconductor industry Semiconductor private equity activities.

business models are Year Deal Price tag

being threatened by 1997 Citicorp Venture and Credit Suisse buy Fairchild from National Semiconductor. US$550 million
1997 Texas Pacific Group buys Zilog. US$527 million
seven disruptive trends
1999 Citicorp Venture and Credit Suisse buy the semiconductor division of Harris. US$520 million in cash and a
in particular: higher Renames the company Intersil. promissory note of US$90 million
costs, competitive 1999 TPG buys the semiconductor components group of Motorola. Renames it ON US$1.6 billion
Semiconductor.
landscape changes,
2004 CitiGroup Venture Capital, Francisco Partners and CVC Asia Pacific buy part of US$828 million
technology convergence Hynix and rename it MagnaChip Semiconductor.
and greater global 2004 Francisco and TPG buy Smart Modular Technologies from Solectron. US$100 million
connectedness. 2005 Kohlberg Kravis Roberts & Co. (KKR) and Silver lake Partners buy Agilient US$2.7 billion
Technologies’ semiconductor unit renaming it Avago Technologies.
2006 Bain Capital buys Texas Instruments’ sensors-and-controls business. US$3 billion
2006 Consortium including KKR and Silver Lake Partners buys majority stake in US$10 billion
Philips Semiconductor. Renames it NXP Semiconductors.
2006 Consortium including Blackstone Group and TPG buys Freescale US$17.6 billion
Semiconductor.

Source: Harbert, Tam. “Private equity chips away at semiconductor industry: Investors see stable cash flows and need for consolidation.”
Electronic Business. December 1, 2006. http://www.edn.com/article/CA6395971.html

Convergence: Semiconductors are the ingly difficult for semiconductor suppliers to
common “fabric” binding digital content find new high-volume semiconductor applica-
Convergence happens when common tech- tion that they need to offset the exponentially
nology creates growing overlaps to create new increasing costs associated with keeping up
consumer value. Convergence and perva- with technological advancement.
siveness are now redefining the market, as
boundaries among application categories are Connectedness: A flatter world brings
dissolving. This blurs the lines of chip applica- people closer
tion and business offering mixes. It seems that nearly everything today is
becoming interconnected. From the social
One area in which convergence trend has networking and user-created content of Web
been most visible is with mobile handsets 2.0, to the ever-increasing number of mobile
increasingly adopting the capabilities of PCs. telephone and Internet users worldwide,
While the product functionality is converging, people want to connect and “talk.” The total
semiconductor suppliers continue to intro- number of worldwide Internet users surpassed
duce more brands, more platforms and more 1 billion in 2005 and is estimated to reach 2
13
feature sets. As a result, it becomes increas- billion by 2011.


Worldwide mobile phone sales are also Ultimately, the increased cost of doing busi-
expected to reach one billion in 2009, with ness will force companies to re-examine their
an estimated 2.6 billion mobile phones in current business models to remain competi-
14
use. Even companies like SanDisk, makers tive. Although the number of suppliers is
of NAND-based flash storage card products increasing drastically – especially within the
that are used in various consumer electronics Indian and Chinese markets – the top 50 semi-
products, are increasing focus on developing conductor supplier companies comprise 84.5
16
mobile platforms. percent of the market.

“[In] the next two to three years Even with so many semiconductor suppliers,
the industry is so capital-intensive that it’s
everything will be wireless, every increasingly difficult to make a significant
kind of devices would be wirelessly impact in the marketplace. As a result, the
connected to everything else and number of new entrants that are significant
enough to be called out by iSuppli has
therefore the distinction between dropped sharply, from 25 in 2002 to just two in
a cell phone or an MP3 player is 2006.
17

going to be somewhat blurred.” For the top 60 semiconductor suppliers, our
– Eli Harari, CEO of SanDisk 15 analysis shows that return on equity ranged
from a low of -67 percent to a high of 46.2
.5
18
When it comes to semiconductors, the race for percent in 2006. Operating profit margin also
“smaller, faster and cheaper” is still on, but the fluctuated widely for this group: from a low of
19
traditional focus on product and technology -14.9 percent to a high of 51.6 percent. During
innovation alone is not sufficient to survive this period, sales growth for this group was
the seven Cs – especially skyrocketing costs, the most volatile measurement of all, ranging
dynamic competition, digital convergence and from a low of -9.5 percent to a high of 102.2
20
greater global connectedness. percent.

Emerging types of industry players As the fabless model gains increased
Companies’ responses to the industry disrup- momentum in the marketplace, the traditional
tions of the seven Cs will ultimately determine integrated device maker (IDM) model remains
whether they win big or lose big. Chip under pressure to transform. Since 2003,
production increasingly resembles a gourmet fabless players achieved higher operating
restaurant kitchen, where numerous chefs profit margin than both IDMs and foundry
line up to add just the right spices to the mix; players; in 2005 and 2006, fabless players
no longer can one company single-handedly doubled IDM and foundry players in return on
bring value to the end user. equity (see Figure 3).


FIGURE 3.
Operating profit margin and return on equity analyses.

30% Operating profit margin analysis Return on equity analysis

20%

10%

0%

-10%
2001 2002 2003 2004 2005 2006 2001 2002 2003 2004 2005 2006

-20%
IDM Fabless Foundry

-30%

Source: IBM Institute for Business Value analysis of Thomson financial data.

Three major roles emerge for IDMs Super Suppliers – This category describes
With the imminent industry shake-out, we suppliers that can continue to “operate its
believe three different types of IDM players will own game” by being a big influencer in the
emerge: Super Suppliers, Alliance All-stars and industry. Today, Intel is the only IDM that fits this
Market Creators (see Figure 4). description. Unlike other IDMs, Intel has the
luxury to buy up value chain players to build
FIGURE 4. and strengthen its vertical integration model.
Three emerging types of IDM players. Through its global investment arm, it can help
shape the greater ecosystem in which it oper-
Self-sufficient and powerful. Super ates.
Super supplier can continue to operate its
Suppliers own game by being a big influencer in Since 1991, Intel Capital has invested US$4
the industry.
billion in more than 1,000 startups in over 30
21
countries. Two recent examples include
Suppliers that form alliances not only
Alliance to create new value proposition in the purchasing London-based social network site
All Stars marketplace, but also to fight the rising Bragster for a reported US$3.5 million and
cost of R&D and manufacturing.
making substantial investment in UK-based
FREEDOM4 Ltd, formerly known as Pipex
Suppliers that focus on core
Market competencies and remain focused on Wireless Ltd, to accelerate Intel’s strategic
differentiating their business offerings mobile WiMAX deployment.
22
Creators
in the marketplace.

Source: IBM Institute for Business Value.


Three types of “We try to make sure that all the Market Creators – These suppliers focus on
core competencies and aim to differentiate
integrated device relevant players in the ecosystem their business offerings in the marketplace.
makers (IDMs)
are ready about the same time.” Like some Alliance All-Stars, Market Creators
are expected to are beginning to embrace the fab-lite strategy
– Arvind Sodhani, president of Intel Capital23
emerge: The Super and to outsource much of their manufacturing.
Suppliers, Alliance Alliance All-Stars – This group consists of
Given that considerable capital investment will
All Stars and suppliers that form alliances, not only to
be needed to move to 32-nm manufacturing
create new value in the marketplace, but also
Market Creators. and beyond, this “asset-light approach” is of
to fight rising R&D and manufacturing costs.
critical strategic importance, especially to
Some Alliance All-Stars are also beginning to
companies with strained balance sheets. As
embrace the fab-lite strategy, a model in which
process technology becomes less differenti-
they no longer maintain their own manufac-
ated, market-specific system know-how – an
turing operations for components with chip
ability to understand the end-product and how
structures above a specific size.
different chips integrate with one another – is
Examples of Alliance All-Stars include fast becoming the true competitive advantage.
Freescale, IBM and Infineon, each of which is
As Market Creators focus more on design
focused on nurturing collaboration with part-
and less on manufacturing, their business
ners as a core competency. When it comes
models will begin to resemble the fabless
to developing new technologies, Alliance
model. Companies like Sony, AMD and Texas
All-Stars would share the costs and risks of
Instruments have publicly announced their
designing new manufacturing processes
fab-lite strategies; however, they have not
by forming alliances. The alliance approach
clearly defined their new business models
allows each partner to later incorporate the
and transition strategies. Only time will tell
co-designed processes into its own manu-
which industry participants will emerge as true
facturing environments. One major benefit
Market Creators.
of this collaborative innovation approach is
that Alliance All-Stars can continue to retain
competency across the entire semiconductor
“It’s really about what the customer
manufacturing value chain without having to wants and what the end customer
invest in the entire manufacturing capability. experience is going to be.”
“You can’t be the leader by yourself – Phil Hester, CTO of AMD 25

anymore. The technology is just too
complicated and expensive.”
– John Kelly, Sr. VP of IBM Research24


Considering the “To fab or not to fab” Recommendations
question With the seven Cs and so much resulting
In the 1990’s, the cost of building a fab reached industry volatility, the proven strategies of the
above the US$1 billion mark.26 Today, it would past may bring failures in the future. For ages,
cost a company US$5 billion or more to build innovation has been a technology-led affair,
300mm wafer fabrication facilities, with additional
with most big breakthroughs coming out of
operation and maintenance costs.27 It is becoming
giant and secretive research labs. It was an
increasingly non-profitable for a company to have
era when big corporations in developed coun-
its own fab, except in cases where a very broad
tries accounted for most R&D spending.
product offering caters to different customer
segments and the economies of scale can justify Our study has shown that higher R&D spent
the fab’s operating costs. doesn’t help ensure better performance in
IDMs must not assume that having a fab is a terms of growth, profitability or shareholder
competitive advantage. In fact, some fabless returns. The game is changing from individual
players have fared well without a fab. Unless IDMs to “ecosystem” competition. No one player
can show that owning fabs will translate into can make it alone. What’s more, in order to tap
higher profitability, they will likely face shareholder into the next growth wave of the industry, the
pressure to pursue fab-lite or fabless strategies. traditional model of focus on technology prod-
When making this decision, some key questions ucts and geographic markets will need to be
for IDMs include: make room for a new business model, which
emphasizes providing the right mix of business
1. Investment. Do you have the investment capacity
offerings to a new set of end-user markets
(US$4 billion or more over two years)?
(see Figure 5).
2. Research. Do you have in-house research
capability to support on-going research in the As companies navigate in today’s rapidly
semiconductor designs, materials and process changing global business environment, their
technology needed to remain competitive? ability to shift direction and introduce business
3. Demand volume. Do you have the volume model innovation is proving to be a critical
necessary to keep the fabs fully utilized? success factor. In the IBM 2006 Global CEO
Study, which was based on interviews with 765

FIGURE 5.
The rules of the game have changed.
Becoming end user-centric
Past Future
Technology product Business offering

Geographic market End-user market



Business model corporate and public sector leaders worldwide, 2) The Enterprise Model approach involves
we found that the financial outperformers put innovation in the structure of the enterprise
innovation can happen
twice as much emphasis on business model and the role it plays in new or existing value
by bringing about
innovation as underperformers. We also found chains. This approach focuses on redefining
change to one or more organizational boundaries.
that business model innovation had a much
of the following: the stronger correlation with operating margin
3) The Revenue Model approach involves
industry model, the growth than other types of innovation.
28

innovation in how companies generate
enterprise model and revenues by reconfiguring offerings (product/
Neither the CEOs we spoke with, nor a review
the revenue model. of the current literature, provided a clear defi- service/value mix) and/or by introducing new
nition of business innovation. Nor did either pricing models. This approach leverages
reveal what type of business model innovation customer choice and preferences, as well as
yields the best results. To find those answers, new technologies.
we conducted a follow-up study that identified
These approaches to business model
three distinct approaches to business model
innovation can either be used alone or in
innovation: via revenue models, enterprise
29 combination.
models and industry models (see Figure 6).
Choosing the right course of business model
1) The Industry Model approach involves
innovation is only one part of the equation.
innovation in the “industry value chain.” This
Business capabilities must also be devel-
can be accomplished via: moving horizontally
oped to optimize the ability and likelihood of
into new industries (for example, Virgin’s focus
succeeding to innovate innovating success-
on superior skills in consumer management),
fully. These capabilities include the abilities to
redefining existing industries (such as Apple’s
center on end-user needs, collaborate and
iTunes) or developing entirely new industries or
integrate (see Figure 7).
industry segments (such as Google and other
search engine companies).

FIGURE 6.
Three ways to enact business model innovation.
Industry model
Solution

Change the way your whole
industry works
Business model innovation Manufacturer Distributor Retailer Consumer
Business offering

Enterprise model
Redefine what you do
Status quo yourself and where you
collaborate
Product

Revenue model A B C D E F

Change the way you charge
Old New your customers
End user market Price



FIGURE 7. Infineon: Becoming a customer-focused
Three key optimization areas. enterprise
Infineon Technologies AG engages in the design,
Integration Collaboration

Solution
development, manufacture and marketing of
semiconductors and system solutions addressing
Business model innovation three central challenges to modern society:

Business offering
energy efficiency, connectivity and security.
In 2003, Infineon set out to transform from a
Status quo traditional product business into a complete
solutions business and focused on acquiring new
Centricity competencies while continue to capitalize on its
Product

competitive advantages.30
Old End user market New To build stronger customer relationships
and become a customer-focused enterprise,
Infineon reorganized itself into smaller, relatively
independent business units. Business units were
Centricity: Strengthen connections with instructed to work closely with systems makers
both customers and end customers in core market areas like cellular to resolve
As the manufacturing of chips becomes technology problems at the R&D level.31 It also
less of a marketplace differentiator, semicon- redefined market segments more specifically to
ductor suppliers, especially IDMs, need to match customer needs. For instance, Wireless
go beyond the manufacturing to understand was divided into Wi-Fi and Bluetooth.
what customers are making and demon- By working closely with its customers, Infineon
strate systems “know-how.” To do so, the has been more successful at translating
go-to-market approach needs to take a more customers’ ideas into actual products, systems
intimate form. The traditional marketing depart- and solutions. As a result, Infineon reached
ment function of conducting marketing studies, number one in power semiconductors, access
establishing demand and developing product products for broadband communication and in
lines to address demand will no longer high-frequency solutions for wireless communi-
work. Instead, companies must engage with cation.32
customers early on during the R&D phase and
incorporate customer input throughout the Collaboration: Aim for radical collaboration
product development lifecycle process. by thinking and acting “big”
To meet the demanding needs of the customer
Cultivating the ability to truly understand
while addressing the escalating financial
what customers and end users are thinking
and intellectual capital needed to remain in
and where the market is going is important;
the business, ecosystem collaboration is no
however, cultivating the ability to transform
longer a luxury, but a necessity. Going it alone
the organization and ways to do business
is not only risky but also impossible, given
based on the understanding of customer, end
that semiconductor suppliers are becoming
user and market needs is the key competitive
more and more dependent on multi-enterprise
advantage. Being a customer-focused enter-
supply chain and fulfillment networks to meet
prise is not about just the company strategy; it
customer needs. Therefore, to meet customer
is primarily about the execution.


Along with selecting the needs and also stay financially viable, many to smart phones like the Apple iPhone –
companies have embraced collaborative inno- companies rely on others in the ecosystem
right course of business
vation – sometimes even collaborating with to integrate new technologies more effec-
model innovation,
potential competitors. tively and provide one seamless solution for
semiconductor companies end customers. Semiconductor suppliers
must hone essential To establish effective collaborative capabilities, can no longer afford to push pre-designed
companies should focus on fully utilizing many
business capabilities, chips to potential buyers. Truly addressing
helpful collaborative tools (for example, Web end customer needs requires a broader,
including the abilities to 2.0 tools) available today to enhance informa- system-focused view – a “holistic design”
center on end-user needs, tion sharing among ecosystems. They should approach. They can no longer restrict
collaborate and integrate. also readjust company strategy and individual themselves to their original core fields of
performance goals to make collaboration a expertise, and they must learn to how to inte-
required part of how business is done. By grate ecosystem style.
combining resources (both financial and intel-
lectual), companies are no longer bounded by Different types of integration can be
their own limitations. The power of many can achieved, based on the level of financial
help meet increasingly challenging industry participation and risk. The level of integra-
demands. tion can range anywhere from adopting an
arm’s-length contract to pursuing M&A that
Integration: Tighten ecosystem integration helps build vertical integration capabilities.
to bring differentiated value to the end Companies should first take a deeper look at
customer customer needs and then determine which
As end-devices become increasingly complex
ecosystem integration approach to under-
– for example, consider the transition from
take.
the first mobile phone (the Motorola “Brick”)

IBM: Pursuing collaborative innovation via an “open ecosystem”
IBM Microelectronics Division is a top maker of application-specific integrated circuits (ASICs) and static
SRAM chips, and a major supplier of chips for communications applications, including wireless phones and
networking equipment. It also offers extensive contract manufacturing, or foundry, services, through which it
manufactures chips for other companies. For IBM, collaborating in an “open ecosystem” was not an option – it
was matter of survival.
In 2003, IBM’s Microelectronics Division had just “lost US$1 billion in 2002 and was on its way to losing
US$252 million in 2003.” 33 Investors wanted IBM out of the microelectronics business; however, IBM viewed
keeping up with leading-edge chip technology as its key competitive advantage in the development of powerful
servers. 34
To keep up with the rising cost of technology innovation development while running a profitable business,
IBM built its “open ecosystem” strategy, also known as Common Platform Alliance. Currently, IBM has an
“open ecosystem” of chip R&D with three Manufacturing Alliance Partners, which include Chartered, IBM and
Samsung, and four Joint Development Partners, which include Freescale, Infineon, STMicroelectronics and
Toshiba.35 This alliance has enabled the IBM division to keep generating cutting-edge technology for its servers
and at the same time, make a profit despite a cyclical downturn in the chip industry.36


NXP: Generating new business through A self-assessment for chip makers
nurturing a struggling ecosystem Industry participants will need to take stock of
NXP, a newly independent semiconductor their current situation in the maturing semicon-
company (founded by Philips), focuses on ductor environment. Thinking through answers
providing semiconductors and associated to the following questions can help companies
software that aims to deliver better sensory
identify ways they can begin to adapt their
experience in the areas of mobile communica-
own business models in order to stay competi-
tions, consumer electronics, security applications
tive amid great change on many fronts.
and others. As part of its growth strategy, NXP
set its eyes on India’s entry-level wireless handset Strengthen connections with both
market (such as handsets that are priced between customers and end users
US$60 and US$149), which is the largest wireless • How do you segment your customers in
handset market segment in India, accounting for order to understand their specific needs?
40 percent of the total Indian wireless demand.37
• How much user-based research does your
To win the market share battle in this highly company currently incorporate into product
competitive market, NXP is working to boost the
design and development? Why and how
Indian manufacturing ecosystem for wireless
does this process need improvement?
handsets and products for new applications, such
as near-field communications, mobile TVs, point- • To what extent does your company’s
of-sale terminals, GPS navigation systems, mobile product influence a customer’s buying
payment and fixed-wireless terminals. decision?

Lacking a venture capital arm, it began investing Aim for radical collaboration by thinking
in companies that will set up electronics and acting “big”
manufacturing units in India and also are • How does your company view information
potential customers for its semiconductors. In sharing outside the organization, and what
addition, NXP hopes to boost the weak Indian
must you do to make it more feasible and
manufacturing ecosystem by working with the
more likely?
Indian Banks Association and others to develop
mobile-payment systems. It has actively engaged • To what extent does your company have
some Taiwanese ODMs (“Original Design formal or informal relationships with adjacent
Manufacturers”) by providing Indian market value chain partners? In which areas
information, helping them in field trials and can you see the greatest need for tighter
validation in the country, and even hosting their connections and communication?
travel to India.38 • How can you build the support among your
company’s employees – across all levels of
the business – that is necessary to enact
“collaborative innovation?”


In the maturing industry, Tighten ecosystem integration to bring
differentiated value to the end customer
companies should
• How much of the value chain does your
assess their current company perform directly, versus relying on
situation in order to look others?
beyond the expected • How well does your company understand
benefits of Moore’s the economics of each part of its opera-
Law, and plan to make tions?
innovative business • What is your company’s plan to foster a
model changes that strong reputation for working fairly with
can optimize their business partners?

capabilities. Conclusion
The effects of the seven Cs are forcing
semiconductor players to move away from
traditional business models and re-evaluate
what industry roles can offer the greatest
returns. The old race of smaller, faster, cheaper
every two years alone will not ensure success
any more. To achieve sustained competitive-
ness, they’ll need to look beyond the expected
benefits of complying with Moore’s Law, and
make innovative business model changes that
can optimize their capabilities, particularly in
the areas of customer centricity, collaboration
and integration.


Authors Contributors
George Bailey was recently appointed General Ken Englund, Partner, IBM Global Services –
Manager of the IBM Microelectronics Division. Americas
In this role, he has responsibility for IBM’s Karan Kapoor, Consultant, IBM Global
semiconductor business. Previously, he was Services – India
IBM General Manager and Global Leader
Steven Kind, Business Solution Professional,
of the IBM electronics industry organiza-
IBM Global Services – Americas
tion. He previously led a worldwide network
of nearly 5,000 IBM employees focused on Waishan Leung, Senior Managing Consultant,
electronics industry clients, and the develop- IBM Institute for Business Value
ment of business and technology solutions to Michael Maslack, Executive Consultant, IBM
meet industry needs. George has appeared Systems and Technology Group
on American television and radio business
Stephen Pierce, Business Solution
programs, including CNBC TV’s “Power
Professional, IBM Global Services – Americas
Lunch,” and has been frequently quoted in
The Wall Street Journal, Fortune, Industry Deepak Sharma, Managing Consultant, IBM
Week, and other management publica- Global Services – Americas
tions and newspapers. He is the author of Hideto Wakui, Partner, IBM Global Services –
Irresistible! Markets, Models, and Meta-Value Japan
in Consumer Electronics and A Thousand
Tribes: How Technology Unites People in Great About IBM Global Business Services
Companies. George can be contacted at With business experts in more than 170
george.bailey@us.ibm.com. countries, IBM Global Business Services
provides clients with deep business process
Wendy Huang is a Managing Consultant within and industry expertise across 17 industries,
IBM Global Services in the Americas, special- using innovation to identify, create and deliver
izing in human capital management consulting value faster. We draw on the full breadth of IBM
within the industrial sector. Wendy has more capabilities, standing behind our advice to
than 10 years of business consulting experi- help clients implement solutions designed to
ence and has worked with clients across the deliver business outcomes with far-reaching
electronics industry in a wide variety of busi- impact and sustainable results.
ness transformation efforts. Wendy can be
reached at wendy.w.huang@us.ibm.com.


References 10
Pausa, Ed. “China’s Impact on the
1
Wang, David T “IEDM 2005: Selected
. Semiconductor Industry, 2006 Update.”
Coverage.” Real World Technologies. http:// PricewaterhouseCoopers. 2007 http://www.
.
www.realworldtech.com/page.cfm?ArticleID pwc.com/techforecast/pdfs/ChinaSemis06_
=RWT123005001504&p=2 web-X.pdf
11
2
Borst, Chris. “Advancing Nanoscience “Should India gamble on microchips?”
through R&D Consortia.” March 25, NanoChina. February 28, 2007 http://www.
.
2008. http://www.nanotech-now.com/ nanochina.cn/english/index.php?option=co
columns/?article=183 ntent&task=view&id=743&Itemid=182.
12
3
“Gartner pulls down its chip industry Harbert, Tam. “Private equity chips away at
forecast.” EE Times India. January semiconductor industry: Investors see stable
16, 2007 http://www.eetindia.co.in/
. cash flows and need for consolidation.”
ART_8800497850_1800007_NT_0be531e6. Electronic Business. December 1, 2006.
HTM http://www.edn.com/article/CA6395971.html
13
4
“Moore’s law made real by Intel innovation.” Internet World Statistics “Internet Growth
Intel Corporation. http://www.intel.com/tech- Statistics: Today’s road to eCommerce and
nology/mooreslaw/?iid=search global trade.” http://www.internetworldstats.
5 com/emarketing.htm
LaPedus, Mark. “Schemes strip cost out of
14
chip R&D.” EE Times. July 16, 2007 http://
. “Gartner Says Mobile Phone Sales Will
www.eetimes.com/showArticle.jhtml;jsession Exceed One Billion in 2009.” Gartner press
id=XZW3CAYR5HSLCQSNDLRSKHSCJUNN release, July 2005. http://www.gartner.com/
2JVN?articleID=201001339 press_releases/asset_132473_11.html
15
6
Hilkes, Rob. “Under the Hood: Uncovering Cooper, Charles and Tom Krazit.
hidden chip costs.” EE Times. October 22, “Newsmaker: SanDisk CEO flashing forward
2007 http://www.eetimes.com/showArticle.
. to phones.” C-NET News. August 31, 2007.
jhtml?articleID=202402983 http://www.news.com/SanDisk-CEO-flashing-
7 forward-to-phones/2008-1041_3-6205436.
Ibid.
html
8
“Next Generation Lithography Technologies 16
IBM Institute for Business Value analysis of
for the Production of Advanced
iSuppli data.
Semiconductors.” Nanotechnology News.
17
Azonano. August 28, 2007 http://www.
. Ibid.
azonano.com/news.asp?newsID=4823 18
IBM Institute for Business Value analysis of
9
“Gartner Says 40 Percent of Today’s Thomson Financial data.
Semiconductor Vendors Likely to Leave 19
Ibid.
the Industry Within 10 years.” Gartner 20
Ibid.
press release. September 13, 2004.
http://www.gartner.com/press_releases/
asset_104967_11.html.


21 28
Roberts, Bill. “Show me the technology.” “Expanding the Innovation Horizon: The
Electronic Business. 8/1/2006. http://www. Global CEO Study 2006.” IBM Corporation.
edn.com/article/CA6355852.html http://www.ibm.com/bcs/ceostudy
22 29
Mutschler, Ann Steffora. “Intel funds Giesen, Edward, Saul J. Berman, Ragna Bell
social network site Bragster.” Electronic and Amy Blitz. “Paths to Success: Three
News. February 13, 2008. http:// ways to innovate your business model.” IBM
www.edn.com/article/CA6532167 . Institute for Business Value. June 2007 http://
.
html?nid=3351&rid=1918692984 www-935.ibm.com/services/us/index.wss/
23
Helft, Miguel. “Venture Investing as ibvstudy/gbs/a1028552?cntxt=a1000401
30
a Strategy, Not to Make Money.” The “Focus on Customers: Expansion of
New York Times. September 22, 2006. solutions business.” Infineon Technologies.
http://www.nytimes.com/2006/09/22/ “http://www.infineon.com/boerse/jahresber-
business/22venture.html?_r=1&n=Top/ icht2003/english/2_1_customer_focus.htm
Reference/Times%20Topics/People/H/ 31
Merritt, Rick. “Chip makers must shift from
Helft,%20Miguel&oref=slogin fabs to systems.” EE Times. January 15,
24
Hamm, Steve. “Big Blue Goes for the Big 2008. http://www.eetimes.com/showArticle.
Win.” BusinessWeek. February 28, 2008. jhtml?articleID=205800635
http://www.businessweek.com/magazine/ 32
“Infineon Technologies at the Annual
content/08_10/b4074063309405.htm General Meeting 2007: Headway Made in
25
Sperling, Ed. “Ecosystem vs. IDM: Is AMD Realignment - Successful Reorganization
onto something?” Electronic News. March of Unprofitable Operations - Concretized
20, 2007 http://www.edn.com/article/
. Earnings Target of 10 Percent.” Infineon press
CA6429394.html?industryid=47037 release. http://www.presseagentur.com/
26
Sidener, Jonathan. “Who needs fabs infineon/detail.php?pr_id=1244&lang=en
33
anyway?: Chip developers have found Hamm, Steve. “Radical Collaboration.”
success without manufacturing capa- BusinessWeek. August 30, 2007 http://
.
bilities.” SignOnSandiego.com. February www.businessweek.com/innovate/content/
8, 2008. http://sports.uniontrib.com/news/ aug2007/id20070830_258824.htm
business/20080208-9999-1b8fabless.html 34
Ibid.
27
“IDM firms go for ‘fab lite.” EE Times Asia. 35
Common Platform: Joint Development
March 16, 2007 http://www.eetasia.com/
. Partners. http://www.commonplatform.com/
ART_8800456825_480200_NT_92ccf68e. about/joint_development_partners.asp
HTM


36
Hamm, Steve. “Radical Collaboration.”
BusinessWeek. August 30, 2007 http://
.
www.businessweek.com/innovate/content/
aug2007/id20070830_258824.htm
37
Krishnadas, K.C. “NXP working to boost
Indian manufacturing ecosystem.” EE Times.
March 7 2008. http://www.eetimes.com/
,
showArticle.jhtml;jsessionid=LMRG2SFYHQ
3H4QSNDLPSKHSCJUNN2JVN?articleID=2
06902451
38
Ibid.


© Copyright IBM Corporation 2008

IBM Global Services
Route 100
Somers, NY 10589
U.S.A.

Produced in the United States of America
05-08
All Rights Reserved

IBM and the IBM logo are trademarks or
registered trademarks of International Business
Machines Corporation in the United States,
other countries, or both.

Other company, product and service names
may be trademarks or service marks of others.

References in this publication to IBM products
and services do not imply that IBM intends to
make them available in all countries in which
IBM operates.

GBE03058-USEN-01

Semiconductor Industry - Needs more than Moore's Law

Recomendados

Recomendados

Más contenido relacionado

La actualidad más candente

La actualidad más candente (7)

Destacado

Destacado (20)

Similar a Semiconductor Industry - Needs more than Moore's Law

Similar a Semiconductor Industry - Needs more than Moore's Law (20)

Más de IBMElectronics

Más de IBMElectronics (10)

Último

Último (20)

Semiconductor Industry - Needs more than Moore's Law