This document summarizes a presentation given by Lucas van Grinsven of ASML about keeping innovation moving. It discusses:
1) ASML is a cornerstone of the chip industry and its strategy is to be a technology leader in lithographic systems, enabling customers to increase chip functionality while reducing cost and power consumption.
2) Sustainability is a key driver of the chip industry as Moore's Law allows chips to double in power every 1.5-2 years while reducing price, leading to growth in applications like smartphones. ASML enables this through lithography equipment.
3) ASML is taking steps to reduce the energy usage and carbon emissions of its business, focusing on reducing emissions per transistor,
Keeping Innovation Moving with Sustainable Semiconductor Technology
1. PublicSlide 1 |
Keeping innovation moving
Lucas van Grinsven
Hoofd Communicatie
“Mobiliteit beweegt iedereen”, 17 november 2010
2. PublicSlide 2 |
Agenda
• Introducing ASML – a cornerstone of the chip industry
• Sustainability is a key driver of the chip industry
• Keeping innovation moving in the Netherlands
4. PublicSlide 4 |
Our strategy
•“
•To be a technology leader in lithographic systems and
software for semiconductor manufacturing,
•thus enabling our customers to increase the functionality
of microchips while reducing the cost and power consumption
per function on a chip
5. PublicSlide 5 |
$300.3 B
Semiconductor
Chips in 2010
$228.4 B in 2009
$1,360 B Electronic Applications in 2010
$1,204 B in 2009
$6.1 B
Semiconductor
Litho market in 2010
$2.5 B in 2009
Source: Gartner Q3/10 and ASML
From silicon to smart electronics
9. PublicSlide 9 |
ASML enables Moore’s law by
providing lithography equipment
to produce smaller and more
powerful chips
The Semiconductor
Manufacturing Process
A variety of
complementary
suppliers provide
the other tools, materials
and packaging equipment
necessary to make ICs
10. PublicSlide 10 |
Moore’s law holds steady for more than 40 years
• Double the computing
power per chip
• At equivalent power
consumption
• For half the price
• Every 1.5 to 2 years
11. PublicSlide 11 |Slide 11 |
Moore’s law: what it means for consumers
Note: data iSupply, March 2009. High quality Flash
0
1
10
100
1000
10000
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
$2,305 for 1Gigabyte (GB)
$/GByte
$0.17 for 1 GB
12. PublicSlide 12 |
Smaller and cheaper chips mean market growth
example: NAND Flash memory
Source: ASML MCC, WSTS, Gartner
1 GB
USB
stick
4 GB
Digital
cameras
8 GB
MP3
player
10 - 20 GB
60 - 80 GB
Hybrid
HDD
FLASH
camcorders
Solid state
disk-based laptops
2 -16 GB
80 - 150 GB
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
’95 ’96 ’97 ’98 ’99 ’00 ’01 ’02 ’03 ’04 ’05 ’06 ’07 ’08 ’09 ’10
Year
FLASHICMarket(millionsofunits)
FLASH units forecast
Several new NAND-based
applications on the horizon
3G
smart-
phones
13. PublicSlide 13 |
Game changer of affordable advanced computing
Source: Morgan Stanley,
The Mobile Internet Report, Dec 2009
10
1000
100
10,000
100,000
1,000,000
0
1960 1970 1980 1990 2000 2010 2020
Devices/Users(MMinLogScale)
1
Minicomputers
10MM+ Units
PC
100MM+ Units
Desktop
Internet
1B+ Units /
Users
Mobile
Internet
10B+
Users???
Mainframes
1MM+ Units
Computing growth drivers over time, 1960 – 2020E
More than
Just phones
• Smartphone
• Kindle
• Tablet
• MP3
• Cell phone / PDA
• Car Electronics
GPS, ABS, A/V
• Mobile Video
• Home
entertainment
• Games
• Wireless home
appliances
Increasing integration
14. PublicSlide 14 |
Source: Jonathan Koomey, Lawrence
Berkeley National Laboratory and Stanford
University, 2009
Moore’s law helps to reduce energy usage
Computations per Kilowatt hour double every 1.5 years
Dell Optiplex GXI
486/25 and 486/33 Desktops
IBM PC-AT
IBM PC-XT
Commodore 64
DEC PDP-11/20
Cray 1 supercomputer
IBM PC
SDS 920
Univac I
Eniac
EDVAC
Univac II
Univac III (transistors)
Regression results:
N = 76
Adjusted R-squared = 0.983
Comps/kWh = exp(0.440243 x year – 849.259)
Average doubling time (1946 to 2009) = 1.57 years
IBM PS/2E + Sun SS1000
Gateway P3. 733 MHz
Dell Dimension 2400
SiCortex SC5832
2008 + 2009 laptops
1.E+16
1.E+15
1.E+14
1.E+13
1.E+12
1.E+11
1.E+10
1.E+09
1.E+08
1.E+07
1.E+06
1.E+05
1.E+04
1.E+03
1.E+02
1.E+01
1.E+00
ComputationsperkWh
1940 1950 1960 1970 1980 1990 2000 2010
15. PublicSlide 15 |
Sustainable Innovations, enabled by ASML
LED-lighting Green datacenters Social & Mobile media
Solar Cells
High-performing & energy efficient Chips
ASML-machines that realize shrink
16. PublicSlide 16 |
Communication became ~ 1013
more energy efficient
enabled by scaling of semiconductors
Frederic Remington, “The Smoke signal”, 1905, Amon Carter
Museum, Forth Worth, USA
5 MJ/b
20 wood sticks of 2 cm diameter and 50 cm
long equals ~3 dm³ Message size 10
characters or 10 ~15 MJ/dm³ energy from
burning wood we use 45 MJ/message or 5
MJ/b
High Speed Downlink Packet Access, HSDPA
speed 3.65 Mb/s using 5.5 W resulting in ~1μJ/b
(Siemens UR5 router)
1 μJ/b
18. PublicSlide 18 |
Managing energy efficiency, a 5-level approach
1. Energy consumption per transistor
2. Energy use of our sites
3. Energy consumption per machine
4. Energy for lithography of one wafer
5. Energy to produce chips
19. PublicSlide 19 |
Energy consumption: Biggest impact is energy use
transistors produced with ASML machines
0
5
10
15
20
25
30
35
ktonnes CO2
CO2 emissions
to produce one
machine
CO2 emissions
one machine
operating one year
CO2 emissions all
transistors
produced per
machine in one
year
20. PublicSlide 20 |
Energy consumption per transistor
CO2 emissions per computational bit decreases every year
• 2007: 42 (103
kg CO2/Petabyte memory)
• 2008: 21 (103
kg CO2/Petabyte memory)
• 2009: 16 (103
kg CO2/Petabyte memory)
The trend of reducing CO2-emissions per computational bit will continue in the future
22. PublicSlide 22 |
Managing energy efficiency, a 5-level approach
1. Energy consumption per transistor
2. Energy use of our sites
3. Energy consumption per machine
4. Energy for lithography of one wafer
5. Energy to produce chips
23. PublicSlide 23 |
CO2 emissions sites 2015 based on production estimates
(Business as usual scenario without CO2-saving measures)
0.0
20.0
40.0
60.0
80.0
100.0
120.0
140.0
160.0
2009 2010 2011 2012 2013 2014 2015
CS w w
ACE
Temp
Rich
Wil
VH
Target: < 45 kton CO2
in 2015
Employee mobility:
<5 kton CO2
26. PublicSlide 26 |
Mobility survey @ ASML
• Survey held in June 2010
• 2277 (of 3931) employees filled out the survey (fix and
flex employees); 58% respons.
• Most important results:
• Only 25% commutes by bike (62% lives within a radius of 15 km)
• Only 4% commutes by bus; public transport is experiences as
inefficient (buss stop is at ASML doorstep)
• 59% commutes by car
• 86% is flexible in start and end time of the working day
• 60% commutes during rush hour
• 13% of employees think that car pooling is an attractive way to
commute (9% is car pooler at the moment)
27. PublicSlide 27 |
Issues and potential for improvement
Issues:
• Commuter packages are not up to date
• Frequent requests by employees for full reimbursement of the
costs for public transport
• Rempte working is facilitated but not promoted
• Car pooling is not facilitated
• Cycling is not actively encouraged
• A new fiscal law will be implemented (1.4% regulation); policy
choices have to be made e.g. a higher amount of electric bikes
Potential for improvement:
• There is potential to increase the use of bike or public transport
and car pool activities
• Alternative ways for commuting are large; potential for change in
behavior is large, interest in car pooling is large
28. PublicSlide 28 |
What is required?
ASML’s Vision on Mobility & Campus:
‘’ASML strives the Veldhoven campus to
be a sustainable place where people can
work and meet, in a productive, safe and
pleasant way, supported by:
• free choice in ways of commuting
• a flexible work environment (in ups and downs)
• stimulus for employees to choose a healthy and
sustainable way of commuting
29. PublicSlide 29 |
Mobility actions
Short term actions:
1. Cycling plan (extension of current plan)
2. Encourage public transport (extension current plan)
3. Promote and support car pooling
4. Green lease car policy (company car park is very small anyway)
Mid term actions (under investigation):
1. Comprehensive employee mobility plan
2. Remote and Flexible work places (“Nieuwe Werken’’)
The electronics world is very large. In 2008 a total of nearly1,550 billion US dollars in electronic applications have been sold. This ranges from computers, camera’s, mp3 players to Blu-ray players and industrial electronics.
Nearly all electronic equipment contains chips. That total market was 262 billion dollars. For the production of chips many different production systems are used. The lithography equipment ASML makes is only one of them, but it is the most critical step and the systems are also relatively expensive. The total market for lithography equipment to make chips was 5.5 billion dollars of which ASML had a share of more than 65% according to various research firms.
This whole system we call the ‘food chain’ of the electronics industry.
For 2009 the situation is much different because of the economic crisis. The data presented is from Gartner and gives data that is always delayed.
The reduction in sales of electronics goods is quite dramatic, but the deeper you get into the foodchain the larger the reductions in revenues are. The same happens when the market is growing: also then the deeper you get in the foodchain the larger the gains are.
Conclusion: the deeper you get into the foodchain of the electronics industry, the larger the swings are.
We are a growing market: market share is growing within an expanding market
Proof of the pudding locations
We make the systems that make it able to follow Moore’s law
Moore’s law means: see bullets and has been proven to be correct in the last 40 years
For you and me when we’re in the Media Market, Moore’s law means that we can get more gigs for less and less money
And one thing leads to another: the smaller and cheaper the chips are, new applications are possible, requiring more and more memory, leading to market growth
By helping reducing the size and cost of memory, we also help reduce the energy needed to calculate.