Packaging the Monolith - PHP Tek 2024 (Breaking it down one bite at a time)
Keynote on industrial internet
1. The industrial internet
B E N E D I C T E V A N S
@ b e n e d i c t e v a n s
O c t o b e r 2 0 1 4
2. Moore’s Law, again
10,000,000,000
100,000,000
1,000,000
10,000
100
1
Transistors per CPU
1970 1980 1990 2000 2010 2020
3. In the last iPhone launch weekend, Apple sold 25x more
CPU transistors than were in all the PCs on Earth in 1995
ALL PCS ON EARTH IN
1995
250m PCs in use
3.2m transistors per
Pentium CPU
<800 trillion
transistors
IPHONE 6 LAUNCH
WEEKEND
10m phones sold
2bn transistors per
A8 CPU
20,000 trillion
transistors
4. Seymour Cray, 1960 – 37
cents per transistor (retail)
Intel 8080 chip, 1974 – 8
cents per transistor
Today, 8 cents for 1m transistors
7. Our grandparents could
count their electric motors
Our parents could count
the things they owned
with a chip inside
We can count our
connected devices
11. ‘Big Data’ brings unprecedented scale
Digital
Information
Created
Source: IDC's Digital Universe Study, April 2014
Challenge: From controlled to chaos at 40% CAGR
1 Exabyte = 1 billion Gigabytes
12. Completing the computerization of industry
Office
PCs &
networks
ERP
‘Industrial
Internet’
A generational change as big as PCs, networks or enterprise software
13. New technology in business has a feedback loop
The
business
shapes the
PRODUCTS CUSTOMERS
tools
The tools
reshape
the
business
14. Finding the tools
CONTROL THE FIREHOSE
Raw sensor data is a
firehose
Need tools to
manage, abstract, co-ordinate,
compress
INGEST THE DATA
How to get this back
into enterprise IT
How to change
enterprise IT -
analytics, Hadoop,
etc, etc
DO SOMETHING USEFUL
Convert to actionable
intelligence
How to drive real-time
execution and build a
feed-back loop
15. The progression of data
‘50s – ‘00s
1.0
Descriptive
2.0
Predictive
3.0
Prescriptive
Mid ’00s – Today
Future
16. Big data transforms business operations
90% prescriptive and predictive analysis
Machine learning closes the feedback loop
Analytics data + streaming big data + machine data
Cycle time: Millions of insights per second
Data Intelligence for all across the org chart
Analytical toolset at the point of decision making
17. The tools can reshape the business
EFFICIENCY
“No more step-ladders”
– data
collection
Operational efficiency
Asset utilization,
maintenance
PRODUCT DESIGN
Knowing exactly what
a product did allows
new designs
Knowing what it’s
doing NOW allows
new designs
SYSTEM DESIGN
Understanding
everything in a
system for the first
time allows a new
system
Continuous feedback
18. Past analogy: containerization
EFFICIENCY
Faster port turn-arounds
Fewer breakages &
theft
Higher fleet utilization
PRODUCT DESIGN
Much bigger, more
efficient ships
New ports, new
onshore logistics
SYSTEM DESIGN
Move to a global
supply chain
19. A computer should not ask anything that it should know
A person should not work out something a computer could
What a computer knows is changing radically