MIT OpenCourseWare
http://ocw.mit.edu
15.912 Technology Strategy
Fall 2008
For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.
Intel Photolithography and
Effective Organization of R&D
Professor Jason Davis
MIT Sloan School of Management
Intel Photolithography Update (1)
• Wilson develops the Consortium option:
– Resolves Free Rider problem with a complicated business model:
“Virtual National Lab” VNL organized:
• EUV LLC company manages research; holds all IP
• Lawrence Livermore + Sandia + Lawrence Berkeley labs run it
– ~150‐170 people working on EUV
• Intel (12‐15p; $50m); Motorola (6‐8p; $10m); AMD (2p;$5m)
– Bought shares in EUV LLC
• Equipment manufacturers (Nikon and Canon) purchased rights of first
refusal with guaranteed margins to convince them to make
photolithography equipment.
Intel Photolithography Update (2)
• Problems emerge for EUV LLC:
– Congress steps in to limit foreign companies value capture…EUV
LLC forced to create incentives for domestic companies
– By 2006, technical progress on 45‐nm EUV faces major problems
including power issues.
– Intel pushes EUV back from 2008 (now!) to 2011 on
Roadmaps…extends optical technologies with “computational”
tricks to correct errors that are smaller than wavelength of light.
Value Creation: Technology, Markets,
and Organizations
How will we
Create value?
How will we How will we
Deliver value? Capture value?
The S‐curve Maps Major Transitions
Maturity
Performance
Discontinuity
Takeoff
Ferment
Time
Summary Novartis Pharmaceuticals: Organizational design reflects research topic synergies
Centralizing R&D permits Novartis to focus resources on common, underlying research topics. The Novartis
corporate senior leadership has traditionally been very technologically oriented (PhD/MD) and drives the
company to develop new blockbusters. Novartis acquires product rights from biotech firms to plug gaps in
their offering.
C D
Technology • Academia leads fundamental research in understanding disease mechanism
• Industry labs identify therapeutic targets based on basic science of disease mechanism
• Patents of composition provide strong competitive insulation for a single molecule product
Situation
Market and • Range of easily exploitable business opportunities is narrowing while the range of radical
Strategy new approaches (e.g. genomics) is expanding rapidly
• Pharma industry is driven by blockbuster products anticipated by street visibility through a
C D 10 year pipeline
• Novartis portfolio decision-making bodies include business unit representation
Organizational • Need high critical mass of researchers to advance the common scientific platforms supportin
Context multiple product categories
C D Early stage product development teams are highly cross-functional
Leadership • Corporate leadership is highly technical and closely involved with the research programs
and development projects.
C D
• Fully centralized • Novartis and other major pharma companies
CEO
research that supports 7- spend 15-20% of sales on R&D to discover
10 therapeutic area and develop “blockbuster” products
Structure
Novartis Global
• Centralized R&D provides critical mass of
businesses R&D Relevance specialized expertise and facilities to leverage
• Scientific platforms science across multiple product categories
Oncology Cardiovascular
support multiple • To offset the potential disconnect between
therapeutic areas (e.g. R&D and business, Novartis deploys research
immunology) C review including senior business,
D
development and manufacturing
representatives.
Intel: Decentralized R&D structure reflects availability of externally‐available
Summary
technology plus a strategic lock on industry standards
Intel’s strategy is distinct among semiconductor makers: Intel exploits externally-led basic research, and
concentrates on adapting it to get products to market quickly. Intel “free rides” on fundamental research from
competitors such as such as Lucent and IBM. They set the rules of the game through their control of
standards and architecture.
Technology • Major technology advancements are acquired from external sources
• Product life cycles are very short
• As current generation technology approaches theoretical performance limits, search intensifies
for radical new approaches
C •D Technology is modular, but Intel has a strategic lock on architectural control
Market and • Semiconductor industry is highly competitive and mature
Situation
Strategy • Intel competes on its ability to develop processes to bring products to market faster than
competitors; flexible production lines allow very efficient variations on existing technology
C • Relies on industry consortium (Sematech) to advance manufacturing state of the art
D
• Emphasis on building strong brand equity to project product “premiumness”
Organizational • High level of acquisition activity
Context • Strong ability to exploit both internal and external research
C •D Places paramount importance on time-to-market issues associated with technology transfer
Leadership • History of strong technological leadership from CEO (from DRAM chips to processors)
C D
• Business-unit
Intel Executive
Decentralized R&D Office • Intel relies on in-sourcing available
• Technology Council technology through an Decentralized
Structure
coordinates research Microprocessor Intenet Communications Technology Relevance R&D structure.
Product Group Group Manufacturing Group
• Architecture Labs • Intel is seldom first to invent, but are fast
conduct basic research Microcomputer Intel Architecture Components Research and efficient adapters
to maintain dominance Research Labs Labs Labs
• This approach has yielded substantial
in setting industry
standards and forecast C
D
profitability, but leaves them vulnerable to
future competitive technological advances.
Current “best practice” attempts to
balance & integrate the two poles:
“Basic”
or “fundamental” Internal or External?
science
EXPLORATION
“Applied”
research
EXPLOITATION
Current “best practice” attempts to
balance & integrate the two poles:
"Research”
Spending $ IBM 1985
Intel, 1985 IBM 1995
Intel, 1985
External Internal
Pharmaceuticals 1981-1997
Pharmaceutical Firms: Relevant Patents vs Basic Research Publications, 1981-1997
Pharmaceutical firms: Relevant patents vs basic research publications, 1981-1997
2500
Merck
2000
Hoechst/MMD
No. of relevant patents
Patents
Bayer
1500
Ciba-Geigy
Eli-Lilly
Pfizer Warner-Lambert
1000
Abbott Hoffman-LaRoche
DuPont
Rhone-Poulenc B-M-S
500
Burroughs-Wellcome
0
0 500 1000 1500 2000 2500 3000 3500
No. of published articles in relevant basic journals
Image by MIT OpenCourseWare.
Publications
Source: Kwanghui Lim: “The Relationship between Research and Innovation in the Semiconductor and
Pharmaceutical Industries: Implications for Theories of Knowledge Spillovers “
Semiconductors 1981-1997
Semiconductor firms: Relevant patents vs basic research publications, 1981-97
4000
IBM
3500
Motorola
Toshiba
Patents
3000 Mitsubishi
No. of relevant patents
Tl Hitachi, NEC
2500
Fujitsu
2000
ST Micro Philips
1500 Siemens
AT&T
Natl semi
1000 Matsushita
Intel
500
NTT
0
0 1000 2000 3000 4000 5000 6000 7000
No. of published articles in relevant basic journals
Image by MIT OpenCourseWare.
Publications
Source: Kwanghui Lim: “The Relationship between Research and Innovation in the Semiconductor and
Pharmaceutical Industries: Implications for Theories of Knowledge Spillovers “
Intel Summary and Take‐Aways
• Creating Value Often Requires Industry‐wide
Coordination
– Technology Leaders can benefit by being first on S‐curves and market
diffusion curves
– Free Rider Problems can emerge
• Tradeoffs between Value Creation and
Value Capture can emerge
• Effective Organization Balances Exploration
and Exploitation
Looking Forward:
• S‐curve papers handed back; most groups did
well.
– Grading focused on ways to create and destroy
value.
• Reflections on Effective Organization:
– Organization Structures:
• Amount of Structure
• Connectivity
• Internal vs. External
– Organization Processes:
• Co‐evolving
• Exploration and Exploitation
• Patching
– Incentives:
• High Power “entrepreneurial” incentives vs. Coordination
