People, Planet, Clouds

A presentation at LJC Unconference in November 2019 in London, UK by Holly Cummins

Slide 1

peop l e, pl a n e t , c lo u d s Holly Cummins IBM Garage @holly_cummins

Slide 2

#IBMGarage @holly_cummins

Slide 3

#IBMGarage @holly_cummins

Slide 4

#IBMGarage @holly_cummins

Slide 5

20 to 60 code changes are submitted every minute #IBMGarage @holly_cummins

Slide 6

40,000 changes a day that’s a lot of builds

Slide 7

1 May 2019: UK government declares a climate emergency

Slide 8

Slide 9

July 2019

Slide 10

July 2019 warmest month

Slide 11

July 2019 warmest month ever

Slide 12

July 2019 warmest month ever worldwide

Slide 13

1° hotter than pre-industrial #IBMGarage @holly_cummins

Slide 14

#IBMGarage Flood by Mauro Lucchesi from the Noun@holly_cummins Project

Slide 15

island nations disappearing photo: Department of Foreign Affairs and Trade underwater fins by Marco Livolsi from the Noun Project #IBMGarage @holly_cummins

Slide 16

30 football pitches destroyed per minute Matt Zimmerman, flickr Axe by jokokerto from the Noun Project #IBMGarage @holly_cummins

Slide 17

#IBMGarage photo: NaveenNkadalaveni @holly_cummins Fire by ✦ Shmidt Sergey ✦ from the Noun Project

Slide 18

#IBMGarage photo: NaveenNkadalaveni @holly_cummins Fire by ✦ Shmidt Sergey ✦ from the Noun Project

Slide 19

USA: 5x more fires 5x longer #IBMGarage photo: NaveenNkadalaveni @holly_cummins Fire by ✦ Shmidt Sergey ✦ from the Noun Project

Slide 20

extinctions Bramble Cay melomys first mammal to be extinct as a result of climate change #IBMGarage photo: Greg Schechter Death by Icon Island from the Noun @holly_cummins Project

Slide 21

“What happens now and in these next few years will profoundly affect the next few thousand years.” - Sir David Attenborough #IBMGarage @holly_cummins

Slide 22

where do we fit in? #IBMGarage @holly_cummins

Slide 23

developer by Adrien Coquet from the Noun Project

Slide 24

impact at work developer by Adrien Coquet from the Noun Project

Slide 25

impact on the world #IBMGarage @holly_cummins

Slide 26

impact on the world (but not that kind of impact) #IBMGarage Meteorite Impact by Ben Davis from the Noun Project @holly_cummins

Slide 27

#IBMGarage @holly_cummins

Slide 28

2017 US data centres used 90 billion kWh #IBMGarage @holly_cummins

Slide 29

2017 45,000,000,000 x x 1 year #IBMGarage Phone by Andrei Yushchenko from the Noun Project @holly_cummins

Slide 30

416 terawatts consumed by datacenters #IBMGarage @holly_cummins

Slide 31

3% of all electricity is used by datacenters #IBMGarage @holly_cummins

Slide 32

1.4 x global data centre energy consumption UK energy consumption #IBMGarage UK by Richard Slade from the Noun Project @holly_cummins

Slide 33

2005: data centres so badly organised only 40% of cool air reached computers #IBMGarage Air Conditioner by Eugene Dobrik from the Noun Project @holly_cummins

Slide 34

60% of cooling energy wasted #IBMGarage Garbage by David from the Noun Project Air Conditioner by Eugene Dobrik from the Noun Project @holly_cummins

Slide 35

#IBMGarage @holly_cummins

Slide 36

80% improvement in data center energy efficiency • low-power chips • solid state drives (instead of spinning hard drives) #IBMGarage disk drive by Idwar from the Noun Project @holly_cummins

Slide 37

Slide 38

IBM z15 #IBMGarage Dinosaur by Nick Bluth from the Noun Project @holly_cummins

Slide 39

IBM z15 #IBMGarage Dinosaur by Nick Bluth from the Noun Project @holly_cummins

Slide 40

IBM zSeries 50% power consumption of x86 (30% more performance) #IBMGarage @holly_cummins

Slide 41

#IBMGarage @holly_cummins

Slide 42

cloud is more energy-efficient than a small data centre #IBMGarage @holly_cummins

Slide 43

is it just data centers? #IBMGarage @holly_cummins

Slide 44

bitcoin #IBMGarage @holly_cummins

Slide 45

== bitcoin #IBMGarage carbon footprint of Denmark @holly_cummins

Slide 46

1 bitcoin transaction == 761,118 VISA transactions #IBMGarage @holly_cummins

Slide 47

1 bitcoin transaction == 50,741 hours of watching Youtube #IBMGarage @holly_cummins

Slide 48

what can we do?

Slide 49

#IBMGarage @holly_cummins

Slide 50

what can we do?

Slide 51

#IBMGarage @holly_cummins

Slide 52

communicate

Slide 53

#IBMGarage @holly_cummins

Slide 54

“Maybe the most important thing you can do on climate might be the simplest: talk about it now.” #IBMGarage @holly_cummins

Slide 55

#IBMGarage (we don’t think there’s much link between climate change and earthquakes - but look at the visualisation!) @holly_cummins

Slide 56

#IBMGarage (we don’t think there’s much link between climate change and earthquakes - but look at the visualisation!) @holly_cummins

Slide 57

#IBMGarage @holly_cummins

Slide 58

understand Map by SANTRI ICON from the Noun Project

Slide 59

models can be improved granularity policy simulation rare events

Slide 60

this is complex machine learning can help we need to innovate techniques #IBMGarage @holly_cummins

Slide 61

#IBMGarage @holly_cummins

Slide 62

common client science bottlenecks: #IBMGarage @holly_cummins

Slide 63

common client science bottlenecks: limited access to weather data #IBMGarage @holly_cummins

Slide 64

common client science bottlenecks: limited access to weather data insufficient computing power #IBMGarage @holly_cummins

Slide 65

common client science bottlenecks: limited access to weather data insufficient computing power insufficient data storage capacity #IBMGarage @holly_cummins

Slide 66

#IBMGarage @holly_cummins

Slide 67

toleration Umbrella by Rockicon from the Noun Project

Slide 68

#IBMGarage @holly_cummins

Slide 69

developers can make situations better #IBMGarage @holly_cummins

Slide 70

cloud democratises access to data #IBMGarage @holly_cummins

Slide 71

cloud democratises access to information #IBMGarage @holly_cummins

Slide 72

when infrastructure goes, the cloud goes OWL Call For Code 2018 winner $200,000 IBM grant

Slide 73

#IBMGarage @holly_cummins

Slide 74

IBM GRAF finer grained predictions updated 6 - 12 x more often #IBMGarage @holly_cummins

Slide 75

climate change has overturned centuries-old wisdom about weather patterns #IBMGarage Plant by Icon Island from the Noun Project Cow by Symbolon from the Noun Project @holly_cummins

Slide 76

best access to forecasts: a weekly newspaper report Newspaper by Loïc Poivet from the Noun Project #IBMGarage @holly_cummins

Slide 77

granular forecasts covering the whole world (not just the rich parts of the world) #IBMGarage @holly_cummins

Slide 78

#IBMGarage @holly_cummins

Slide 79

change

Slide 80

#IBMGarage @holly_cummins

Slide 81

I don’t want your hope. I don’t want you to be hopeful. I want you to feel the fear I feel every day. And then I want you to act. I want you to act as if the house is on fire. Because it is. fire house by Andi Nur Abdillah from the Noun Project #IBMGarage @holly_cummins

Slide 82

“products with sustainability claims generally outperformed the growth rate of total products in their respective categories.” #IBMGarage @holly_cummins

Slide 83

efficiency #IBMGarage @holly_cummins

Slide 84

#IBMGarage @holly_cummins

Slide 85

#IBMGarage @holly_cummins

Slide 86

release many times a day automate everything #IBMGarage @holly_cummins

Slide 87

do we need to stop continuous integration and deployment? #IBMGarage @holly_cummins

Slide 88

how do you build a monorepo with 9,000,000 source files? #IBMGarage @holly_cummins

Slide 89

how do you build a monorepo with 9,000,000 source files? you get very good at dependency graphs #IBMGarage @holly_cummins

Slide 90

#IBMGarage @holly_cummins

Slide 91

optimise for fast feedback (side effect: efficient builds) #IBMGarage @holly_cummins

Slide 92

Hey boss, I created a Kubernetes cluster. IBM Garage @holly_cummins

Slide 93

Hey boss, I created a Kubernetes cluster. I forgot it for 2 months. IBM Garage @holly_cummins

Slide 94

Hey boss, I created a Kubernetes cluster. I forgot it for 2 months. … and it’s £1000 a month. IBM Garage @holly_cummins

Slide 95

IBM Garage @holly_cummins

Slide 96

QUARKUS supersonic subatomic Java

Slide 97

quarkus is light

Slide 98

… still light with CRUD

Slide 99

… and quarkus is fast

Slide 100

TrueNorth chip neurosynaptic lower power use

Slide 101

what we imagine when we widen roads

Slide 102

what we get

Slide 103

Slide 104

8% of CO2 emissions one industry: cement #IBMGarage @holly_cummins

Slide 105

unintended consequences #IBMGarage @holly_cummins

Slide 106

is my travel sustainable? #IBMGarage @holly_cummins

Slide 107

is my travel sustainable? #IBMGarage @holly_cummins

Slide 108

is my travel sustainable? #IBMGarage @holly_cummins

Slide 109

#IBMGarage @holly_cummins

Slide 110

static string concatenator() { string result = “”; for (int i = 0; i < 314159; i++) { result += getStringData(i); } return result; } #IBMGarage @holly_cummins

Slide 111

<div class=”time”>stuff</div> <div class=”gravatar32”>stuff</div> <div class=”details”>stuff<br/>stuff</div> #IBMGarage @holly_cummins

Slide 112

string s =”<div class=”time”>” + st() + st() + “</div> <div class=”gravatar32”>” + st() + “</div> <div class=”details”>” + st() + “<br/>” + st() + “</div>”; return s; #IBMGarage @holly_cummins

Slide 113

var sb = new StringBuilder(256); sb.Append(“<div class=”time”>”); sb.Append(st()); sb.Append(st()); sb.Append(“</div><div class=”gravatar32”>”); sb.Append(st()); sb.Append(@”</div><div class=”details”>”); sb.Append(st()); sb.Append(“<br/>”); sb.Append(st()); sb.Append(“</div>”); return sb.ToString(); #IBMGarage @holly_cummins

Slide 114

100,000 iterations dual core 3.5 GHz Core 2 Duo #IBMGarage @holly_cummins

Slide 115

100,000 iterations dual core 3.5 GHz Core 2 Duo #IBMGarage Simple Concatenation 606 ms String.Format 665 ms string.Concat 587 ms string.Replace 979 ms StringBuilder 588 ms @holly_cummins

Slide 116

“the minute you begin worrying about tiny little optimizations, you’ve already gone down the wrong path” -Jeff Atwood #IBMGarage @holly_cummins

Slide 117

every improvement is good, but … #IBMGarage Theater by Symbolon from the Noun Project @holly_cummins

Slide 118

knowledge helps us focus on optimisations that matter

Slide 119

a laptop uses $8 a year (yes, you are allowed to run your unit tests) #IBMGarage Plug by johartcamp from the Noun Project @holly_cummins

Slide 120

how can we know the eco impact? #IBMGarage @holly_cummins

Slide 121

#IBMGarage @holly_cummins

Slide 122

waste == bad waste == cost cost == bad #IBMGarage @holly_cummins

Slide 123

but it’s complicated #IBMGarage @holly_cummins

Slide 124

Sheffield Shenfield #IBMGarage @holly_cummins

Slide 125

Sheffield Shenfield #IBMGarage @holly_cummins

Slide 126

Sheffield £47.00 Shenfield #IBMGarage @holly_cummins

Slide 127

Sheffield £47.00 Berlin Shenfield #IBMGarage @holly_cummins

Slide 128

Sheffield £47.00 Shenfield #IBMGarage £21.37 Berlin @holly_cummins

Slide 129

Sheffield £48.80 Shenfield #IBMGarage £37.57 Berlin @holly_cummins

Slide 130

borrowing against the future #IBMGarage @holly_cummins

Slide 131

borrowing against the future oh … and infrastructure #IBMGarage @holly_cummins

Slide 132

example what is most sustainable to eat? #IBMGarage @holly_cummins

Slide 133

#IBMGarage Plant by Icon Island from the Noun Project Meat by The Icon Tree from the Noun Project Chicken by Виталий Плут from the Noun Project Fire by ✦ Shmidt Sergey ✦ from the Noun Project @holly_cummins

Slide 134

tofu #IBMGarage meat Plant by Icon Island from the Noun Project Meat by The Icon Tree from the Noun Project Chicken by Виталий Плут from the Noun Project Fire by ✦ Shmidt Sergey ✦ from the Noun Project @holly_cummins

Slide 135

#IBMGarage tofu meat GM soya chicken Plant by Icon Island from the Noun Project Meat by The Icon Tree from the Noun Project Chicken by Виталий Плут from the Noun Project Fire by ✦ Shmidt Sergey ✦ from the Noun Project @holly_cummins

Slide 136

#IBMGarage tofu meat GM soya chicken imported local Plant by Icon Island from the Noun Project Meat by The Icon Tree from the Noun Project Chicken by Виталий Плут from the Noun Project Fire by ✦ Shmidt Sergey ✦ from the Noun Project @holly_cummins

Slide 137

tofu meat GM soya chicken imported local deforestation #IBMGarage Plant by Icon Island from the Noun Project Meat by The Icon Tree from the Noun Project Chicken by Виталий Плут from the Noun Project Fire by ✦ Shmidt Sergey ✦ from the Noun Project @holly_cummins

Slide 138

#IBMGarage tofu meat GM soya chicken imported local deforestation fed imported soy Plant by Icon Island from the Noun Project Meat by The Icon Tree from the Noun Project Chicken by Виталий Плут from the Noun Project Fire by ✦ Shmidt Sergey ✦ from the Noun Project @holly_cummins

Slide 139

complex trade-offs #IBMGarage @holly_cummins

Slide 140

do less #IBMGarage @holly_cummins

Slide 141

reduce waste #IBMGarage cutting back Smile, Stress by David from the Noun Project @holly_cummins

Slide 142

Slide 143

Slide 144

no, remote meetings don’t match face to face

Slide 145

#IBMGarage @holly_cummins

Slide 146

“The situation we’re facing isn’t how to do more with less. It’s how to do less with less.” #IBMGarage @holly_cummins

Slide 147

the future

Slide 148

“A zero-carbon world doesn’t have to be an awful one. The more we invest in it, the more the infrastructure around it changes to support it, and the easier and cheaper it gets… and the more normal it gets.” - Liz Keogh (@lunivore) #IBMGarage @holly_cummins

Slide 149

Talk about it Encourage your organisation to do more Lower your footprint Use your skills #IBMGarage @holly_cummins