Meet IBM Research

Dr. Matthias Kaiserswerth
Director and Vice President, IBM Research - Zurich

Meet IBM Research

1 © 2012 IBM Corporation

IBM Research - Zurich

Agenda

Introduction & Strategy
How an idea becomes an innovation at IBM

Dilemmas of Strategic Innovation Management

Questions & Discussion



Agenda




IBM Research: The World is Our Lab



IBM Research: The World is Our Lab

• World's largest information technology
research organization
• Patent Leadership
• More than 3,000 scientists and engineers
• IBM spent $6.3B on R&D in 2011
• 5 Nobel Prize Laureates



IBM Research -Zürich

Computer Science: Security and Privacy,
Business Integration Technologies,
Distributed Computing, Systems Management

Systems: Server Technology, Accelerator
Technology, I/O Link Technologies,
System Software, Energy Management

Storage: Tape Technologies,
- Founded in 1956 Heads and Actuators, Storage Systems,
Memory and Probe Technologies
- 45 Nationalities

- 25% students, pre- & post-docs, visiting Math & Computational Sciences:
scientists Business Optimization,
Computational Sciences, Data Analytics
- 90 Collaborations with Partners
Science & Technology: Semiconductors,
- Two Nobel Prizes Systems, Beyond the Transistor,
(1986 and 1987) Nanotechnology, Biotechnology

- Recently opened the Binnig and Rohrer
4 Nanotechnology Center with ETH Zürich © 2012 IBM Corporation


Expanded Goals for our 2nd Century

2012 and beyond
Lead
–IBM to new technologies/businesses
–Stronger IBM/client collaborations
–Deeper industry expertise

“The World is our Lab”
–Enable our new labs to succeed
–Focus on growth markets
–Real world research

Even Bolder Leadership Agenda
–Leadership offerings
–A new era of computing
–Far-reaching research and grand challenges



IBM Research and Development
IBM continues to make consistent and sizable investments in R&D

$5.8B
6180
Patents
R&D Investment

6% E/R 6% E/R

~2950
Patents

2005 2011



IBM Research: Supporting IBM’s Growth

Growth Markets
22% of IBM’s
geographic revenue

Growth
Initiatives
60%

Cloud Business Analytics
16% revenue increase
3.8 times 2010 revenue from 2010

Base
Research
40%

Smarter Planet
Nearly 50% revenue
increase from 2010
7


IBM’s Growth Market Expansion
In 2011 we opened nearly 100 new branch offices

8


The World is Our Lab

Dublin

Zurich
Watson China
Almaden
Haifa Tokyo
Austin India

Kenya
Brazil

Melbourne

IBM Research labs
Labs added since 2010

9


Africa: The Next Frontier for IBM Research Innovation

New lab in Kenya will conduct basic and
applied research to solve East Africa’s
problems, such as transportation
congestion and to optimize the use,
storage, safety and distribution of water.

10


Agenda






How an Idea Becomes an Innovation at IBM

Technology
Trends Industry Solution
Labs

Industry Acquisitions
Trends
Collaboration
Prototype
Idea

Trend Scan
3-10 years out Pilot
Research Deployment /
Social Productization
FOAK
Trends

Licensing
Academic
Patents Papers Research
Solution Centers Transfer
Experience Papers

Ideas management
12 Business © 2012 IBM Corporation
management


Trend Scans



Trend Scans

Global
Innovation
Outlook
 Dialog about innovation,
business transformation
and societal progress
 Collaborate across a global
ecosystem
 GIO 4.0: security and
society, water and oceans



Trend Scans

jam
Global
Innovation Jamming
Outlook
 Dialog about innovation,  Collaboration between
business transformation companies,
organizations and
 Collaborate across a global family members to
ecosystem
yield first-of-a-kind
 GIO 4.0: security and opportunities
society, water and oceans



Trend Scans

jam
Global Global
Innovation Jamming Technology
Outlook Outlook
 Dialog about innovation,  Collaboration between
business transformation  Identify emerging
companies, technology trends
organizations and significant to industry in
 Collaborate across a global family members to
ecosystem the next 3 - 7 years
yield first-of-a-kind
 GIO 4.0: security and  Has a direct influence
opportunities
society, water and oceans on IBM’s technical
strategy


2012 Global Technology Outlook

Managing
Uncertain Systems of People
Data at
Scale
The Future Watson

Outcome Based Business

Future of Resilient Business and Services
Analytics




Technology
Labs

Trends
Collaboration
Prototype
Idea

Trend Scan
FOAK
Trends

Licensing
Academic
Solution Centers Transfer
Experience Papers

Ideas management


Idea/Concept

se
IBM Press Relea
rs
sio ns in data cente
Intelligent use of heat emis
of the future IBM's
air scientists from
eBIT 2008 tradef on
2008 — At the C ept of a zero carb
erland, 3 March a pioneering conc -
Rüschlikon, Switz for the first time extremely energy
nover, Germany, ory will present otype is not only s
Han esearch Laborat at. The first prot at buildings. Thi
(NYSE : IBM) Zurich R the direct reuse of waste he redirected to such uses as to he stem
nter achieved by ing energy to be water cooling sy
emission data ce rs of the IT operat sts. A power ful new kind of
ws three-quarte all operating co
efficient, but allo and lowers over
lly friendly .
is environmenta of this innovation
e chip is the basis
embedded on th industry.
allenge to the IT
time, it poses a ch al
e IT industry urce. At the same r about 2% of glob
Challenges in th ngly precious reso dustry accounts fo d Lawrence Berke
ley
sable yet increasi communications
in
Ene rgy is an indispen ts, the information and fic the world over. The US-base llion KWh,
tner Group analys nerated by air traf nts to some 120 bi
According to Gar is equal to that ge data centers amou nters
carbon dioxide em
issions, which gy consumption
of operate the data ce
th e worldwide ener atts each. Thi s energy is used to
ory confirms that ts of 1000 megaw
National Laborat pical power plan y.
roughly equivale
nt to 14 ty our service econom total
nstitute the backbone of and 50% of their
around the world that co tly spend between 25 But
Companies curren heating systems.
cent ers is increasing. ped into ov ersized electrical aste heat is
nsumption of data nters have develo at. Instead, this w
And the energy co ns. Indeed, data ce rect supply of he
center operatio to serve as a di ed.
IT budget on data annel this energy e not being realiz
not be en possible to ch that huge po tential savings ar
until now, it has nment, meaning
simply dissipat ed into the enviro annover,
08 tradefair in H
esearch - Zurich is . At the CeBIT 20 eves a
tion from IBM R want to change th ta center that achi
chnological solu or y in Zurich now energy-efficient da g
Te
Research Laborat ncept of a highly energy and coolin
Scie ntists at the IBM first time the pioneering co s have design ed an innovative most —
ill present for the at. The researcher ing is needed the
Germany, they w using its waste he plied where cool
int by directly re water-cooling sy
stem that is ap
zero carbon footpr is a new kind of
of which
circuit, the basis g buildings, swim
ming
ctly on the chip. the chip for heatin ent for
dire heat generated by important requirem
er model is to reuse H owever, the most al
mission data cent heating network. r modern municip
The goal of the zero-e g the heat into a municipal e a certain th reshold, which, fo
ply for conductin waste heat be abov
pools etc. or sim mperature of the
heat is that the te
the direct use of e has an
about 50 °C. than air. This alon
he ating networks, is 4000 times better the cooling
can conduct heat e able to improve
ers used water, which g on the chip , researchers wer
is, IBM research ating water coolin
To accomplish th tential. But integr
-saving po
enormous energy of magnitude.
16 efficiency by several orders © 2012 IBM Corporation


Collaboration and (Working) Prototype

d
to build new kin
se
IBM Press Relea
plan
IBM and ET H Zurich unveil r
of water-cooled supercompute
ons by up to 85% issi
carbon-dioxide em
energy consum ption by 40% and ute of
heat. Aims to cut iss Federal Instit
Dir ect reuse of waste -aware co mputing, the Sw ater-cooled
to achieve energy first-of-a-kind w
09—In an effort d plans to build a , dubbed
and, June 23, 20 ) today announce innovative system
Zurich, Switzerl M (NYSE: IBM ty buildings. The to 30 tons
ich (ETH), and
IB for the universi mated to save up
Technology Zur repu rpose excess heat system by up to 85% and esti
that will directly footprint of the
supercomputer ease the carbon nologies.1
ct ed to decr stem using to day’s cooling tech
Aquasar, is expe d to a similar sy to 50% percent of
an
of CO2 pe r year, compare rtaking. In fact, up ing
a staggering unde ting but by power
ta centers energy-efficient is n today is no t caused by compu hen looking
g systems and da ergy consumptio r from optimal w
Making computin on footprint or en situation that is fa
data center ’s carb ocessors from ov
erheating—a
average air-cooled ems to keep the pr
ing syst
the necessary cool a holistic perspect
ive.
cannot afford anym
ore to
at ener gy efficiency from facing in the 21st century. We ns Prof. Dr.
humanity will be ce alone,” explai
e num ber one challenge nal sp eed and performan lead investigator
of
“Ene rgy is arguably th the criteri on of computatio in Emerging Technologies and rcomputers
systems based on Thermodynamics consumption supe
design computer the Laboratory of ormance an d low net power
H Zurich, head of must be high perf
Poulikakos of ET t. “The new target
ary projec
this interdisciplin oling.” ter, which will be
located at
s means liquid co e new supercompu
and data centers. Thi stem is
use, Aquasar—th n by 40%. The sy
and direct heat re ergy consumptio ell
er-cooling system reduce overall en ater-cooling, as w
With an innovative wat art oper ation in 2010, will ntists in the fie ld of chip-level w urich Lab.
d is planned to st TH and IBM scie ntists at IBM’s Z
the ETH Zurich an collaboration of E advanced by scie
m joint research ergy re-use”
based on long-ter ers with direct en
“water -cooled data cent and will have a pe
ak
as on a concept for ladeCenter® servers in one rack
ist of two IBM B
superc omputer will cons
The water-cooled input and
about 10 Teraflops
.2 ocessor, as well as
perf ormance of uid cooler per pr entire
a microscale high
-performance liq cted easily to the
with to be connec ted and disconne
will be equipped allow each blade
Each of the blades nnections, which
tworks and co
output pipeline ne ransporting
e). air, and its heat-t
system (see imag es m ore efficiently than fficient to keep th
e
capture he at about 4,000 tim re of approx imately 60°C is su ant results
t has the ability to a water temperatu rature of the cool
Water as a coolan level cooling with high input tempe
far superior. Chip- e maximally allo
wed 85°C. The
properties are also res well below th .
ratu will be about 65°C
chip at operating tempe output, which in this case
-grade heat as an
in an even higher


Deployment

se
IBM Press Relea

ed SuperMUC
Hot-Water Cool percomputer
u
Named Eu rope’s Fastest S
e petaflops system
massive thre
mp uting center gets
German superco 40% less energy
while consuming r the
rcomputer built fo
, an IBM hot- water cooled supe l system in
oday, SuperMUC the most powerfu
many, June 18, 2012: T many, has been named rillion calculations
per
HAMBURG, Ger RZ ) in Garching, Ger ta flops (three quad ip hot-
puting Centre (L mance of three pe innovative on ch
Leibniz Supercom maint ain a peak perfor e air-cooled m achine due to its
ly, the system can than a comparabl
Europe. Incredib rcent less energy
nsuming 40 pe
second) while co many.
water cooled desi
gn. in Hamburg, Ger
Supercompu ting Conference
the International
ere repo rted today during UC gets its power
The official results w d Zurich, Sw itzerland, SuperM XXX,000
gen, Germany an work of more than
sc ientists in Boeblin ocessors -- equivalent to the to provide
veloped by IBM ,576 Intel Xeon pr is then repurposed
Designed and de iDataPlex with 18 er (140F), which
from an IBM Syst
em x® ee Celsius wat
ith up to 60 degr
PCs. Each proc essor is cooled w
100%
campus heating. quired to purchase
institutions acro ss Germany are re st in class
ed by state-funded c community a be
electricity consum ving the scientifi an of the
his year all of the mitment, while gi dt Bode, Chairm
“T hel p us keep our com as never befo re,” Prof. Dr. Arn
. SuperMUC will predict outcomes
sustainable energy experiments and
ries, design
system to test theo
rs of LRZ.
Board of Directo ed by
t today is not caus
an d carbon footprin l, but
ling system gy consumption ts choose a radica
Inno vative water-coo cooled data center's ener dress this challe nge, IBM scientis en air.
of an average air- g systems. To ad more efficiently th
Up to 50 percent necessary coolin heat 4,000 times
powering the candidate becaus
e it removes
computing, but by cooling, an ideal directly to
not new concept
- water which are attached
micro-chann el liquid coolers
d on bio-inspired
vative cool ing system is base
SuperMUC's inno generated. case is up to
here most heat is ut, which in this
the processors, w en higher-gra de heat at the outp
er results in an ev
ture of th e transported wat
The high tempera
(149 F).
65 degrees Celsius




Technology
Labs

Trends
Collaboration
Prototype
Idea

Trend Scan
FOAK
Trends

Licensing
Academic
Solution Centers
Experience Papers

Ideas management



Technology
Labs

Trends
Collaboration
Prototype
Idea

Trend Scan
FOAK
Trends

Licensing
Academic
Solution Centers
Transfer
Experience Papers

Ideas management


Technology Transfer

se
IBM Press Relea
s
IBM Research create
world’s smallest 3D map
ctures and
f nanoscale stru
chnique improves development o
New 3D microscopic te devices
of the earth
created a 3D map
(NYSE: IBM ) scientists have a new,
pril 2010—IBM d this by means of
and Sa n Jose, CA, 22 A he scie ntists accomplishe a sharpened penc
il —
Zur ich, Switzerland, on on e grain of salt.1 T — 100,000 times smaller than rning
0 of them could fit with a sharp apex plexity. This patte
so small that 1,00 a tiny, silicon tip ced cost and com logy,
breakthrough tech
nique that uses ers at greatly redu future chip techno
sm all as 15 nanomet s in fields su ch as electronics,
and structures as nanosized object
to create patterns cts for developing
new prospe
technique opens , and opto-electro
nics. fferent materials
for
medic ine, life sciences ral 3D and 2D patterns, using di
team created seve
ue capability, the ials: gh, was
e technique’s uniq d Advanced Mater 8 m (14,692 ft) hi
To demonstrate th ific jo urnals Science an lpine mount ain that soars 4,47
rted in the scient rhorn, a famous A
each one as repo plica of the Matte 00
igh 3D re on.2 er. At this size, 10
• A 25-nanometer-h glass, representing a scale of 1:5 billi micrometers was “written” on a polym roughly eight
ar
created in molecul ly 22 by 11 rrespond to
orld measuring on eters of altitude co in only 2 minutes
and 23
• Complete 3D map of the w salt. In the relie f, one thousand m and was created
d fit on a grain of ring 20 nm2,
world maps coul xels, each measu
compo sed of 500,000 pi chnique for typica
l
nano meters (nm). It is onstrating the viability of the te
into silicon, dem
seconds. ch ed 400-nm-deep
IBM logo was et
• 2D nano-sized applications.
nanofabrication y, very sharp silic
on
e a team of IBM scientists, is a tin
nd the techniqu as developed by
The science behi chnique, which w apex. ing
nt of the new te nanometers at its d tools for produc
The core compone th and only a few uality methods an broad
nometers in leng e existence of high
-q – Zurich. “With its
tip measuring 500 na tim ately linked to th Armin Knoll of IBM Research l tool for
technology are in ains physicist Dr. logy is a powerfu
“Advances in nano on surfaces,” expl tterning methodo
s and objects notip-based pa
nanoscale pattern capability, this na
uniq ue 3D patterning
functionality and scans the
ry small structures
.” that controllably
ge nerating ve be ndable cantilever d
, is attached to a y applying heat an
atomic force microscopes a millionth of a millimeter. B omilling”
the kind used in one nanometer— erating like a “nan
The tip, similar to ith the accuracy of d patterns, thus op
trate material w d on predefine
surface of the subs trate material base
ized tip can remove subs
force, the nano-s
ultra-high precision.
20 achine with
m © 2012 IBM Corporation


Technology Transfer



Agenda






Dilemma #1
Bottom Up and Top Down



Dilemma #2
Long-term research and short-term success

vs.



Dilemma #3
Promotion of intellectual property

26


Dilemma #3
Promotion of intellectual property
6,000

5,000

Number of Patents
4,000

3,000

2,000

1,000

1993 1994 1995
1996 1997 1998
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
0
2011

26


How do you confront these dilemmas?
Here is a snapshot from IBM Research.



For Dilemma #1
High Reputation
Phase 2: Reputation Ranking
Reputation
Management
of Corporate
R&D Centers

Reputation Management of Corporate R&D Centers, Universität St. Gallen, 2009



For Dilemma #1
Strategy Management System
LINE
Lab Director/VP
Direct Report*
Senior Manager

STRATEGY
Strategist
Area Strategist
Interlock
Sub-Strategist

Interlock at the appropriate level within each cell of
Interlock at the appropriate level within each cell of the line X strategy matrix
the line X strategy matrix


For Dilemma #1, #2
Research with
zu Dilemma #1 & #2
Market Success



For Dilemma #1 & #2
Balanced Research Agenda

European Research Council
Volume 9
| Number

bioengine
ering
Advanced Grant 2011
23 | 2009

biology, & ges 3313– 3452
hysics, r 2009 | Pa
emistry, p
mbe
23 | 7 Dece
| Number
tion for ch Volume 9
Miniaturisa 10
OL 2 NO try
R 2010 V is
OCTOBE .com/naturechem
ure
rg/loc
www.nat

Robert-Wichard-Pohl-Preis 2011
www.rsc.o

ACM SIGSAC Outstanding
Innovation Award 2010
Lab on a
Chip

R&D Magazine Top 100 in 2010

Best Innovation European Identity
Award 2010
Lee with SERS
detection
Label-free (2009)9:23
;1-U
Delamarche munodiagnostics 1473-0197
Pages 3313

197
ISSN 1473-0
One-step
im r HIV
Demirci tification fo
sc.org NTHESIS
Automate
d cell quan

TOTAL SYind
Gerhard-Ertl-Preis for Young
Lu t
easuremen
y Number
207890 Pressure m

Four of a k
–3452

ION
N REACT
EDUCTIO g
R
OXYGENrough blockin
Better th TALYSIS
Scientists in 2010
TION CA
R-OXIDA ide stability
WATE es prov
Nanotub
g shot
olecule mu
Single-m 7/9/10 13
:52:16

.indd 1
ve r Oct 2010
nchem co


For Dilemma #3
Open Innovation

Venture
Capitalists ISVs Business
IT Analysts
Partners
Investors Alumni

Regulatory Bodies
Innovation Ecosystem Universities
Clients

Community
Employees Leaders
Standards Competitors
Policy-
Bodies makers



For Dilemma #3
Open Innovation



For Dilemma #3
Innovating with Partners

 10% of our Pre/Post Docs are from
EPFL
 13 CDAs/JDAs with EPFL
 IBM Research Prize (computational
sciences)
 30 EU FP Collaborations in the past
five years

EU Flagship Collaborations Lorem ipsum

2011 Prize was awarded to
Dr. Alessandro Marinoni (right)

The
34


For Dilemma #3
Project STEEPER

Increase electronic device
efficiency by 10× and
eliminate power
consumption of devices in
standby mode

Expertise and research to
tunnel field effect transistors
(TFETs) and semiconducting
nanowires to improve the
efficient use of energy in
electronics

Based on silicon (Si),
silicon-germanium (SiGe)
and III-V semiconducting
nanowires
35


For Dilemma #3
Big Bang Meets Big Data
 Dishes of the Square Kilometer Array will produce
10 times the global internet traffic.

 The SKA computer system will have the processing
power of about one hundred million PCs.

 The technology to handle the SKA doesn’t exist on
the market today



For Dilemma #3
Technology Roadmap with ASTRON

 Green supercomputing: 3-D chip stacks,
high-performance storage systems

 Nanophotonics and optical interconnect
technologies to optimize large data
transfers

 Storage and Memory: Phase change
materials and advanced tape

 Five Year, Public Private Partnership with
Dutch government

 Open Platform for Additional Partners

Meet IBM Research

More Related Content

What's hot (20)

Viewers also liked (10)

Similar to Meet IBM Research (20)

More from IBM Research (9)

Recently uploaded (20)

Meet IBM Research