Title: Knowledge for Development Seminar
1RD Systems at IBM Lessons from Flat-Panel
Display Commercialization James
McGroddy Consultant and former Senior Vice
President of Research, IBM
Knowledge for Development Seminar Center for
International Development Kennedy School of
Government, Harvard University 11 December 2003
2Outline of Talk
- Comments on Innovation Systems in general,
Innovation-supporting Knowledge - The IBM Innovation System - focus on the
front-end - How do you figure out what to
do ? - the management system - some
generalizable principles - Results Some extra-base hits, focus on flat
panel displays - Results Some ground - outs and strikeouts
- Some lessons-learned, some principles
3Innovation Systems
- Complex, tangled, multiply connected, retrograde
.. - Path to success has many stages, many risks, lots
of fog - Kinds of relevant Knowledge - Indigenous -
technical, subject matter - Exogenous -
future market insight, pathfinding skills, etc. - Very different for different segments, even at a
detailed level - agriculture,
healthcare (pharma vs IT ), telecom/IT, the
internet - Generalizing either Theory or Pragmatics a tough
challenge - although some principles and
lessons-learned can be extracted
4Innovation of What vs. Innovation of How
- Many systems can be parsed into Muscle and the
Nervous System - Exponential progress in IT, telecom have
dramatically changed the performance and cost of
the nervous system - Much innovation is the result of applying
upgraded nervous systems to old muscle,
displacing atoms with bits
5IBM RD, focus 1990s
- No major change since then, good results
- R, the front-end of D, and more
- A true Central Research - 600m/yr, 3000
people, 8 labs worldwide - program about 50
self-determined - 50 joint with company
partners ( helps the ) - Charged with Innovation across all segments
6(No Transcript)
7Evolution of IBM Research
through the 70s
invention and technology transfer
the 80s
Joint Programs
direct work with customers
the 90s
the white space
8The Value Chain
Customer Solutions Application Frameworks Horizont
al Applications Platforms .. DB, Transaction,
etc Operating Systems Hardware Systems Processors,
Switches, Routers ICs,Storage, Lasers,
Displays Silicon, Magnetics Electrons, Atoms,
Mathematics
Solutions
Products
Technologies
JC McGroddy
9The Value Chain How Far Can We See ?
Customer Solutions Application Frameworks Horizont
al Applications Platforms .. DB, Transaction,
etc Operating Systems Hardware Systems Processors,
Switches, Routers ICs, Storage, Lasers,
Displays Silicon, Magnetics Electrons, Atoms,
Mathematics
Solutions 1 - 2 yrs
Products 5 - 10 yrs
Technologies 10-20 yrs
gt 50 yrs
JC McGroddy
10Growth of the IT Industry
revenue
2X
Poker
new stuff
more old stuff
X
Chess
old stuff
0
0
5
year
11IBM Business Evolution(remember that this chart
is from 1994 and needs to be scored)
Chess
Poker
12(No Transcript)
13Planning / Execution / Evaluation - Annual Cycle
- The Technology / Business Outlook - 5-10 yr
view - process led by research, inputs from the
business, customers, other external
sources - technology, exploitation, impact
on multiple sectors - shared and debated
internally and externally - backdrop for
strategies - Develop strategy focus - start, stop, grow,
shrink - Prepare proposed plan, too large by 10, rank
projects - lots of involvement of partners - Internal reviews to the top of the business,
closure around budget - Quarterly highlight reviews internally and with
partners - Annual structured assessment, lots of partner
input, overall score - Capture key lessons learned
- Repeat the cycle ..
14Contention and Competition
- The Customer is not Always Right
- Research Investments are made ( partly) to create
Alternatives - The System will often reject Alternatives -
uncertainty, no short term payoff - competition - Alternative Paths may be needed
15Flat Panel Displays . The DTI Story
- Early 80s, lots of candidate technologies -
plasma, EL, LC in different forms, etc. - Outlook and low-end product strategy focused on
centrality of FP displays in 1984 - Internal players/products major CRT-based
monitor group ( Raleigh), plasmagas panel (
Kingston) IBM very vertical, internal - Studies between Research, product group,
identified TFT/LCD as likely winner in 84. Issue
is how, when to pursue. - Initial proposal in 1985 totally internal RD,
raise the mice in the elephants cage - rejected
in favor of partnering with a Japanese consumer
electronics partner. TFT/LCD at about 2 inches
appearing. Trickle Up expected to win. - Small team, high level support (also many
skeptics) chose Toshiba, negotiated joint RD
deal, very foreign to IBM culture - Within 6 months of beginning negotiations, 10/86,
IBM engineers joined the Toshiba team at a new,
first-pass, Toshiba pilot line
16Flat Panel Displays . The DTI Story contd
- 1987 new pilot line completed at IBM, team moves
there - Rapid early successes, 9 inch prototypes, yet
major skepticism within IBM - wrong technology,
wrong partner, blowing the schedule - 1989, despite naysayers, JV mfg company, DTI
started - exploited Japan capital environment - The race is on other vendors such as Sharp,
Hosiden, tool industry developing, Japan focused - New technologies begin to appear, esp. Field
Emission - Rapid progress leads to product commitment
- First IBM Color Laptop announced 3/24/92,
ThinkPads followed - IBM products had bigger,
brighter displays - Ultimately grew to 7B
annual sales - In early-90s shortage, IBM, Toshiba, built
strong market positions - By 2001, Korea major supplier, TV main focus, DTI
sold
17DTI - IBM Initial Motivation
- TFT technology would trickle up
- Consumer Volume Mfg Experience
- The Japanese Way
- Cost of Capital
18DTI - later IBM Motivation
- Rapid Pace of Progress
- Effectiveness of the Relationship
- Commonality of Goals
- Cost of Capital
19IBM
Toshiba
products
products
OEM
IBM Display Business Unit
Toshiba
Liquid Crystal Division
DTI - array, cell - 3 plants -
engineering
IBM DBU development
Toshiba development
IBM Research
Toshiba Research
Display Technology Inc. - an IBM-Toshiba Joint
Venture
20DTI - Lessons Learned
- In a dynamic industry, winning in the new stuff
is key to success. - New stuff often based
on disruptive technology - Bold approach required for disruptive
technologies - poker, not chess -
separation, protection essential - a separate
cage for the mice - Focused partnership with common, not
complementary, goals and be a major help - Often, the technology will almost certainly
develop in the market, while the winners are
determined by management boldness and skill, and
a dose of good luck.
21IBM - Some ground-outs and Strikeouts
- Routers etc NSFnet built with 3 generations of
IBM routers. Threat to SNA kept IBM out of IP. An
alliance might have helped. - PC processors and OS Early equity investment in
Intel not seen as alliance. Proposed investment
in Microsoft rejected in 1986. - WDM and laser groups sold, to Tellabs, JDS
Uniphase
22Inhibitors to Success in New Stuff
- The new opportunity is small and uncertain
- Budgets are fully committed to the current line
- Incremental investment in the current line will
return more - We dont want to confuse our customers
- The new technology is immature
- The new approach competes with our mainstream
approach and threatens a major revenue stream
23Market Sweetspots
All -Optical LAN
Telcos
CIOs - intrasystem
24Three Waves of Computing
Power of choice moves to the end-user
ease-of-use is key
Pervasive Personal Use
Business operators buy the technology, impose it
on the end-users
Customers and users are the High priests
Customers, users, were the high priests
PC and C/S
Mainframe
Server
1960
2020
1980
2000
25Increasing Focus Worldwide
Stepping up the pace New business
models Importance of VC Increased risk
acceptance
Improving the linkage Joint funding/management Mov
ing people across boundaries RD in the
Marketplace
Setting a better agenda Balancing the
portfolio Changing the culture Increasing
external focus Looking toward impacts
Application, Commercialization
Basic, Applied Research Invention
Reduction to Practice