[Music plays and a split circle appears with photos in each half of the circle flashing through of various CSIRO activities and the circle then morphs into the CSIRO logo]
[Image changes to show Dr Sarah Pearce talking to the camera and text appears: Dr Sarah Pearce, Acting Chief Scientist, CSIRO, Deputy Director, Astronomy & Space Science]
Dr Sarah Pearce: Hello everyone. I’m Dr Sarah Pearce. I’m the Acting Chief Scientist at CSIRO and Deputy Director of CSIRO Astronomy and Space Science.
[Images flash through of a rear view of a male walking through a laboratory and then a close view of a researcher putting on gloves, and then two researchers working in a lab]
Welcome to the Generation STEM Masterclass series. You’re in for a treat.
[Images flash through of two researchers in conversation, a digital model on a laptop screen, a close view of a piece of equipment, a female looking at a circuit board, and researchers working in a lab]
Get ready to step outside the classroom, albeit virtually, and enter the fascinating world of science, technology, engineering and maths, or STEM for short.
[Image changes to show Sarah talking to the camera again]
STEM is powering some of the fastest growing industries of our time, industries that are shaping the world we live in today, and the world where you’ll live and work in the future.
[Images flash through of three floors of a cross section of an office building, people walking through an airport, a male looking at an iPhone, highways in a city, and people looking at iPads]
Thanks to the rapid advance of technology there’s been an explosion of new STEM career opportunities for young people like you.
[Image changes to show Sarah talking to the camera again]
Demand for bright, skilled people in STEM is growing almost twice as fast as for other jobs.
[Images flash through to show a researcher looking at a Smart microscope slide, a microchip strapped onto a bee’s back, a 3-D printer, and a road leading into a city lit up at night]
This series will be a window for you into what those industries and jobs will look like and how your interests could lead you to an exciting and rewarding career in STEM.
[Image changes to show Sarah talking to the camera again]
At your age I wanted to be an astronaut and that led me to a career in physics, working in the areas of Space and Astronomy.
[Images flash through of a car moving past telescopes, an aerial view of cars moving through the ASKAP array, a close view of a telescope, and a close view of swirling colours morphing into an eye]
I’ve managed some of the biggest telescopes in the world including one coming soon that will let us look right back to the beginning of the universe. It sounds unbelievable but it’s real and it’s my job.
[Image changes to show Sarah talking to the camera again]
You’ll hear from some of Australia’s most talented experts throughout this series and see how STEM can lead you to a career in sustainability, health, agriculture, aerospace, or any one of a number of other areas.
[Image changes to show a rear view of a male looking out over a city and then the image changes to show a close view and then a profile view of a male looking into a microscope and looking at a screen]
But not everyone needs to be a scientist or engineer.
[Images move through of a female looking up, a researcher looking through a Google glass, a female looking into a microscope, a female wearing a wearable headset, and a close view of a camera lens]
There’s an enormous range of careers in STEM that require a broad variety of skills from communications, to law, to creativity and design.
[Image changes to show Sarah talking to the camera again]
There is something for everyone and I hope you’ll find something for you. Good luck and enjoy the Masterclass.
[Music plays and the CSIRO logo and text appears: CSIRO, Australia’s National Science Agency]
[Images move through to show a rear view of Dr Sophie Calabretto walking towards the Gadi supercomputer, Sophie exiting the data hall and entering another data hall at the supercomputer]
[Image changes to show Sophie talking to the camera and text appears: Dr Sophie Calabretto, Applied Mathematician & Fluid Mechanist]
Dr Sophie Calabretto: Hi. My name is Dr Sophie Calabretto and I’m an Applied Mathematician and Fluid Mechanist. We’ll talk a little bit about what that means later on. Today we will be exploring the fascinating world of Supercomputers.
[Image changes to show a profile view of Sophie talking]
I’ll be talking about how I use Supercomputers in my work and how Supercomputers help scientists all over the globe, tackle some of the biggest problems the world faces today.
[Images move through of a spinning globe, a satellite weather model, cross-section models, a temperature deviation modelling map, and various weather maps and text appears: NCI is supporting researchers at ANU in their efforts to learn more about Earth’s Inner Core]
Supercomputers have over the past 20 years become central to modern science. From climate modelling and weather forecasts to molecular simulations and human genomics, supercomputers help scientists understand the complexity behind some of the most complex and important phenomena of the natural world.
[Image changes to show a facing view of Sophie talking to the camera]
In the coming decades, supercomputers will help us design drugs to treat deadly diseases, build more efficient planes, predict extreme weather events far in advance and much more.
[Image changes to show a profile view of Sophie talking on the right and text appears on the left: In Today’s Lesson, What is a supercomputer?]
Today in this lesson, we will we answer the question “What is a Supercomputer?”.
[Image changes to show a facing and then profile view of Sophie talking to the camera and new text appears on the left: History of supercomputers, We’ll meet Gadi]
We will explore the interesting history of supercomputers and introduce you to Gadi, the most powerful supercomputer in Australia, and talk about the development of the Setonix supercomputer and other supercomputers in Australia.
[Image changes to show a facing view of Sophie talking to the camera]
I’ll discuss my own career path and the importance of being open and curious, when paving your own STEM career.
[Image changes to show a profile view of Sophie talking and new text appears on the slide: How supercomputers can solve real world problems]
We’ll take you on a journey to see how supercomputers solve real world problems and learn some new skills along the way.
[Image changes to show a facing view of Sophie talking to the camera and new text appears on the slide: STEM careers]
Finally, we’ll show you the wide range of STEM careers available to you, if you’re interested in working with supercomputers like me.
[Music plays and text appears on a dark blue screen: What is a supercomputer?]
[Image changes to show a facing view of Sophie talking to the camera and then the image changes to show a blue screen showing text x=100²]
Here is an example of a mathematical problem that you can easily solve.
[Image changes to show a light blue screen and text appears on the screen: (x-100)2.173= 12.3729]
Now consider this problem. You can solve this, but it requires many more steps and probably a calculator.
[Image changes to show a facing view of Sophie talking to the camera and then the image changes to show a long equation on a blue screen]
However, we can solve this easily and in a fraction of a second – it took about 0.1 seconds on my laptop – with any linear algebra application, such as MATLAB, which will run on your laptop or desktop computer.
[Image changes to show another equation on a light blue screen]
Now imagine you have a set of equations that look like this. These equations describe how fluids, such as water flow, and are what I deal with on a daily basis.
[Image changes to show a facing view of Sophie talking to the camera and then the image changes to show a profile view of Sophie talking to the camera]
A linear algebra package like MATLAB cannot solve these in this form, and a regular computer would struggle to do it in a reasonable amount of time.
[Image changes to show a facing view of Sophie talking to the camera]
To solve these equations, we need a very powerful computer to perform a lot of calculations very quickly.
[Image changes to show a profile view of Sophie talking to the camera]
In this situation we would use a supercomputer.
[Image changes to show a view of the data hall in a supercomputer and then the image changes to show a close view of the computer processors]
A supercomputer is made up of thousands of connected computer processors designed to work together at the same time.
[Image changes to show a close view of two processors working together and then the image changes to show a view of the data hall again]
We call this working in parallel.
[Images move through of various very close views of parts of the supercomputer]
A supercomputer can solve large, complex scientific research and modelling problems. For comparison, most of our laptops would only have two or four processors.
[Image changes to show the NCI Australia sign on the side of the supercomputer]
In fact, my laptop can’t even solve that set of equations in the way I want it to.
[Camera pans along the side of the supercomputer]
If I tried, it would crash my laptop every time, which is why I need to use a supercomputer.
[Music plays and text appears on a dark blue screen: Why are supercomputers important?]
[Image changes to show a facing view of Sophie talking to the camera]
Supercomputers are an increasingly important tool for scientists working on some of the biggest questions in science today.
[Image changes to show a spinning world globe and then the camera zooms in and in on the world globe and then images flash through of close view of bubbles, an eye, a brain, and a galaxy]
Some kinds of natural phenomena are so complex and wide-ranging, and require so much data, the only way to truly understand them is to replicate them with the computer models.
[Image changes to show a researcher looking through a microscope, and then the image changes to show a male looking at a piece of equipment]
There’s only so much that we can do with physical experiments.
[Image changes to show a male looking at another piece of equipment, and then the image changes to show a close view of a lens, and then the image changes to show a digital landscape model]
Answering questions about how bushfires spread, what molecules are doing when they interact with cells, and how planes create turbulence as they fly, needs simulations run on a supercomputer.
[Image changes to show a facing view of Sophie talking to the camera]
In the area of health, researchers are studying the COVID-19 virus to understand exactly what it does to the human body and the cells it interacts with.
[Image changes to show a profile view of Sophie talking to the camera]
They are also comparing it to their existing database of medical drugs to see if some existing ones could be useful in responding to the pandemic.
[Images move through to show different digital weather maps and climate model maps, and then the image changes to show Sophie talking to the camera]
Climate change is already impacting us in a big way, and researchers using supercomputers in Australia involved in all aspects of it, climate models about how fast the world is warming, improved extreme weather forecasting, inventing new battery technologies, and making more efficient planes and ships.
[Image changes to show a profile view of Sophie talking to the camera and then the image changes to show a facing view of Sophie talking to the camera]
A supercomputer lets us respond during bushfire emergencies, and then learn from them afterwards as well. Supercomputers allow deep and complex analysis of datasets, simulations and models, which can be referred to as deep learning. Those findings can then be applied to everyday situations.
[Images move through of different views of cars moving along various highways]
For many years, supercomputers were tasked with analysing road and traffic data to undertake deep learning, which has now led to new features in modern cars such as drive assist.
[Image changes to show a computer modelling programme of simulated model cars moving along a road]
Drive assist allows cars to read the road and automatically react accordingly.
[Image changes to show a facing view of Sophie talking to the camera and then the image changes to show a profile view of Sophie talking to the camera]
Think sensors on the windows that pick up the absence of other cars, cars that brake to respond to road conditions.
[Image changes to show a facing view of Sophie talking to the camera]
These technological advancements are driving the future of self-driving cars.
[Music plays and the image changes to show text on a dark blue screen: What are some applications of supercomputers?]
[Images flash through to show a facing view of Sophie talking, a close view of the Covid-19 virus, small particles falling through tessellated wire rolls, and various weather modelling maps]
Some applications of supercomputers include solving questions that benefit us all; world class discoveries; critical response systems for emergencies; and playing a key part in the incremental research and national science priorities.
[Images move through to show various bushfire modelling simulation maps]
Some examples are: Bushfire modelling – The Bureau of Meteorology is developing a specific fire version of their national weather model that can help us respond quicker to fires and learn about how they spread.
[Image changes to show a male working on a laptop and then the image changes to show the laptop screen he is working on]
This has already led to improvements to procedures following an at-the-time unexpected fire event in Western Australia some years ago.
[Image changes to show a facing view of Sophie talking to the camera and then images move through of a close view of simulated models of Covid-19 at work, and drugs defeating Covid-19]
COVID-19 - Understanding the structure of the virus proteins, how they infect human cells, how we might counter them using drugs we already have or develop vaccines or other drugs that can respond effectively.
[Images move through of various atmospheric pollution simulation modelling maps]
Atmospheric pollution simulations – An atmospheric chemistry researcher is looking at urban and industrial pollution and how it spreads and travels through the atmosphere.
[Image changes to show a facing view of Sophie talking to the camera and then the image changes to show a close view of a turbine spinning]
Or, in applied fluid simulations a bunch of people are doing huge fluid simulations around engines and turbines and how to increase efficiency for plane engines and gas turbines.
[Images move through of a close view of a jet engine, a jet moving along a runway, and then Sophie talking to the camera again]
Even a 1% increase in the efficiency of jet engines would save huge amounts of emissions. This includes looking at hydrogen combustion engines as well as future power generation options.
[Music plays and the image changes to show text on a dark blue screen: What is the history of supercomputers?]
[Image changes to show Sophie talking to the camera and then the image changes to show a profile view of Sophie talking]
Supercomputers as we know them today have been around since the 1980s, but technology continues to advance rapidly.
[Image changes to show Sophie talking to the camera and then the image changes to show a profile view of Sophie talking]
A current iPhone is about as powerful as the National Computational Infrastructure’s supercomputer was from 2004.
[Image changes to show a facing view of Sophie talking to the camera]
There are two national scale supercomputers in Australia, and a dozen or so smaller ones.
[Image changes to show the NCI Australia and the Pawsey logos on the screen and then the image changes to show a facing view of Sophie talking to the camera]
NCI and the Pawsey Supercomputing Centre are at the petaFLOP scale. That’s a quadrillion calculations per second.
[Image changes to show the NCI building and then the image changes to show the inside of the building and the camera pans around]
NCI houses the Gadi supercomputer and a huge number of data collections and virtual environments for data analysis and collaboration.
[Image changes to show an aerial view of the Pawsey Centre and then images move through of the inside of the building and then a Pawsey Supercomputing Centre sign on a door]
The Pawsey Centre houses the Magnus and Galaxy supercomputers and hosts the Nimbus cloud, which is specially designed for data-intensive research work in cutting-edge fields such as space science.
[Image changes to show Sophie talking to the camera]
CSIRO houses Pearcey and Ruby and is in the process of developing an even more powerful new supercomputer.
[Image changes to show a rear view of a male walking inside the Pawsey Centre and then the image changes to show a male working on a laptop inside the building]
The Pawsey Centre is developing the Setonix supercomputer which will become the fastest in the Southern Hemisphere in 2022.
[Image changes to show two males looking at the supercomputers and talking and then the camera pans around the room and then the image changes to show a profile view of Sophie talking]
To give you an idea about the plan for the extra power, the existing supercomputer at the Pawsey Centre, Magnus and Galaxy together have 1.83 petaFLOPS of raw compute power.
[Images move through to show a facing view of Sophie talking to the camera, a profile view of Sophie talking to the camera and then a facing view again]
Setonix is forecast to deliver 50 petaFLOPS of power, enough to keep a single person busy calculating for 1.5 billion years just to match what it can do in an instant.
[Images move through of the interior of the Pawsey Supercomputing Centre, a male opening one of the doors of the supercomputer, and then the camera panning up the supercomputer]
The system has been designed to give Australian researchers an edge in emerging research fields such as artificial intelligence and machine learning.
[Music plays and the image changes to show a dark blue screen and text appears: Welcome to Gadi]
[Image changes to show a view of the NCI Centre and the camera pans over the roof of the building and then the image changes to show an aerial view looking down on an oval outside the building]
As I mentioned before, Australia’s current fastest supercomputer Gadi is housed in the NCI Centre.
[Image changes to show a view of the NCI centre and the camera pans in an anticlockwise direction]
The NCI sits on the land of the Ngunnawal people, the Traditional Owners of the Canberra region.
[Image changes to show Sophia talking to the camera and then the image changes to show a profile view of Sophia talking]
NCI supercomputer’s name, Gadi, comes from the Ngunnawal language. It means “to search for”, which is a great representation of the search for knowledge that NCI and all of my fellow researchers are pursuing.
[Image changes to show the Gadi supercomputer inside the NCI building and then the image changes to show Sophie walking around the side of the computer and looking at the painting on it]
What’s more, the Gadi supercomputer’s artwork, is also painted by a local Ngunnawal artist, Lynnice Church.
[Image changes to show Sophie opening the doors of the Gadi supercomputer and entering the data hall]
The two big circles represent the Traditional and the western knowledge systems, coming together guided by the Ngunnawal Elders over many generations.
[Images move through of Sophie walking towards the supercomputer inside the building, and then a male walking around the supercomputer in the building]
Gadi is massive and fast. At the time of launch in 2020, Gadi was the 24th fastest supercomputer in the world.
[Image changes to show Sophie walking down the data hall corridor in the Gadi supercomputer]
It can do 9.3 quadrillion calculations per second.
[Image changes to show the data hall corridor in the supercomputer and then the image changes to show Sophie opening a door on the supercomputer]
It has 180,000 processors.
[Image changes to show a close view of hard drives in the supercomputer]
Remember, your laptop would usually only have two or four.
[Image changes to show a close view of Sophie pulling out the hardware tray in the supercomputer and then the camera pans up to Sophie’s face]
It has 640 GPUs and it can do data transfer at 200 gigabits per second.
[Image changes to show Sophia talking to the camera and then the image changes to show a profile view of Sophia talking]
So how is GADI so fast? Supercomputers work in parallel, sharing operations across hundreds or thousands of processors all accessing a shared pool of data storage.
[Image changes to show Sophia talking to the camera]
You need: Fast networks; Fast processors; Organised data stores and; High-performance code.
[Image changes to show a rear view of Sophie walking down the corridor in the data hall of the supercomputer]
Some of the biggest supercomputing projects are also big users of data.
[Image changes to show a facing view of Sophie walking towards the camera in the data hall of the supercomputer and then the camera follows her around the outside of the supercomputer]
Gadi is so useful for researchers, not just because of the computing power, but because of the integration with more than 50 petabytes – 50,000 terabytes – of research data.
[Image changes to show Sophie talking to the camera]
That includes thousands of satellite pictures of the Earth, climate model data, human genomes, molecular structures and more.
[Image changes to show a profile view of Sophie talking to the camera and then the image changes to show a facing view of Sophie talking to the camera]
For researchers working in environmental science for example, they now have easy access to high-resolution images of Australia that they can use to understand bushfire risk, erosion, water flows, flooding and much more.
[Image changes to show a profile view of Sophie talking to the camera and then the image changes to show a facing view of Sophie talking to the camera]
What does it take to keep a machine like this running?
[Images move through of the corridor of the data hall, inset glass panels in the floor looking into the floor below, and the cooling and electricity systems in the floor below the super computer]
Behind and under the main data hall are all sorts of electricity, cooling and back-up systems that keep all of NCI’s various data and computing systems operating.
[Image changes to show a facing view of Sophie talking to the camera]
There are batteries that kick in automatically in case of a black-out, and diesel generators that can keep hard drives spinning if the black-out lasts more than a few minutes.
[Images move through of chilled water tanks in the building]
There are also large water tanks here that provide a stock of water for cooling.
[Images move through to show chilled water tanks and pipework, a close view of green pipework, and then steam rising above the outside roof]
Water is piped in over the processors to take up the heat they produce, and that hot water is taken off to the roof to evaporate away.
[Images move through of the water vapour above the building]
On a cold winter’s morning, that’s the cloud of water vapour that you can see above the building.
[Image changes to show a facing view of Sophie talking to the camera]
But Gadi is also energy efficient. It is eight times more powerful than the supercomputer that came before it, but only uses twice as much power.
[Images move through to show views of the NCI building and then various views of the solar panels on the building roof]
NCI also has solar panels covering the roof of the building, which along with the ACT’s 100% renewable energy supply, goes a long way to reducing the GADI supercomputer’s environmental impact.
[Image changes to show a facing view of Sophie talking to the camera]
Supercomputers are a critical part of science all around the world. Sometimes they let researchers look in detail at natural phenomena that are impossible to accurately measure or observe.
[Images move through to show various digital maps of the Earth’s surface, tessellated wire roll type designs, and then various coloured shapes moving through the tessellated wire rolls]
At other times, they speed up the research process by working alongside experiments, like with material design and clinical medicines.
[Images move through of various digital maps and then the image changes to show Sophie talking to the camera again]
Big data plays a big role in all these fields of science, and analysing, processing, and storing and sharing all that data is why a national facility like NCI and the GADI supercomputer is so necessary.
[Image changes to show Sophie exiting the area of the GADI supercomputer and then walking towards the camera]
There are thousands of research supercomputers around the world, mostly run by national science infrastructure or partnerships, universities or other research bodies.
[Image changes to show a facing view of Sophie talking to the camera and then the image changes to show a profile view of Sophie talking to the camera]
In the race for the fastest supercomputer in the world there have been some big players such as Japan, the USA, and China.
[Image changes to show a facing view of Sophie talking to the camera]
On a world scale, Australia punches above its weight which is pretty impressive.
[Music plays and text appears on a dark blue screen: The ‘M’ in STEM]
[Image changes to show a facing view of Sophie talking to the camera]
Now I’ll share a little about how I use supercomputers in my work as an applied mathematician.
So firstly, what is an applied mathematician?
[Image changes to show a profile view of Sophie talking to the camera and text appears: The Mathematical Sciences, Pure Maths, Applied Maths, Statistics]
The mathematical sciences are broadly split into three categories: Pure Maths, Applied Maths and Statistics.
[Image changes to show a facing view of Sophie talking to the camera]
Statistics is the collection, analysis, interpretation and presentation of data. Pure Maths is the study of mathematics itself whereas Applied Maths is using maths to solve real-world problems.
[Image changes to show a profile view of Sophie talking to the camera and text appears: Applied Mathematics, Mathematical Biology]
Some of the different branches of Applied Maths are: Mathematical Biology – which is the modelling of biological processes, everything from cancer modelling to neuroscience to migration behaviour in animals to epidemiology.
[Image changes to show a facing view of Sophie talking to the camera and new text appears on the left: Optimisation]
Optimisation is finding the optimal solution which is relevant to a variety of fields such as electrical, civil or control engineering, economics and finance, operations research and industry.
[Image changes to show a profile view of Sophie talking to the camera and new text appears on the left: Dynamical Systems, Material Science]
Dynamical Systems which is the study of things that evolve in time, such as population growth or celestial motion. Material Science is using maths to determine the interesting and important properties of natural and manufactured materials.
[Image changes to show a facing view of Sophie talking to the camera and new text appears on the left: Financial Maths, Fluid Mechanics]
There’s Financial Mathematics and of course there’s Fluid Mechanics.
[Image changes to show a profile view of Sophie talking to the camera and then the image changes to show a facing view of Sophie talking to the camera]
Fluids are everywhere. Air is a fluid, so is water, blood, toothpaste, Jupiter and yet we do not always understand them at a fundamental level.
[Image changes to show a profile view of Sophie talking to the camera]
By that I mean, we can solve certain fluid flow problems. For example, let’s consider laminar flow in a pipe.
[Image changes to show a close view of a blue fluid flowing in a layer through the centre of a pipe of clear fluid]
By laminar I mean that the fluid is nicely behaved and flows in layers.
[Image changes to show a facing view of Sophie talking to the camera and then the image changes to show a profile view of Sophie talking to the camera]
We can write down equations that describe the movement of fluid in a pipe and we can solve those equations that is we can find a solution that will tell us exactly what the fluid is doing anywhere in the pipe at any time in the future.
[Image changes to show a close view of a blue fluid flowing in a layer through the centre of a pipe of clear fluid]
However, we can’t find solutions for a lot of other fluid flow problems.
[Image changes to show a close view of a blue fluid flowing randomly through the centre of a pipe of clear fluid]
For example, turbulent or messy flow in a pipe. But why is that? It’s because turbulent fluid flow is governed by chaotic dynamics and we can’t predict chaos.
[Image changes to show a facing view of Sophie talking to the camera]
Mathematical chaos, in a nutshell, is unpredictability. You may have heard of this in terms of the butterfly effect.
[Image changes to show a profile view of Sophie talking to the camera]
A small change can result in large differences in a later state. For example, a butterfly flapping its wings in Brazil can cause a hurricane in Texas.
[Image changes to show a facing view of Sophie talking to the camera]
This isn’t true but it’s meant to describe the sensitive dependence on initial conditions which is a very important trait of chaotic dynamics.
[Image changes to show a pendulum swinging back on forwards on a black screen]
Consider a single pendulum. A single pendulum is not chaotic. This is because we can write down a straightforward equation that captures the physics of the pendulum system, and we can solve it. At any time in the future, we know exactly where the pendulum is going to be and how fast it will be moving.
[Image shows the pendulum having an extra arm added to it and the image shows the pendulum swinging erratically using the extra arm]
Now imagine that the pendulum has an extra arm attached to the end. Rather than swinging back and forth, it can now trace out a completely erratic path.
[Image changes to show two double pendulums swinging in the identical erratic paths]
However, if I let a double pendulum go from exactly the same spot, it will trace the same crazy path, but it has to be exactly the same.
[Image changes to show a profile view of Sophie talking to the camera]
This is because the double pendulum is deterministic. There is no randomness to the way it moves. If it were random, dropping the pendulum from the same spot would result in a different path.
[Image changes to show the two double pendulums being started from a slightly different position and the image shows two erratic but different pendulum paths being traced]
But if I let the pendulum go again, changing the initial position by just a fraction, what starts off looking like almost the same path, completely diverges and ends up tracing out something that is completely different.
[Image changes to show a profile view of Sophie talking to the camera]
This is an example of chaos. A small change initially can lead to huge unpredictable changes in the future.
[Image changes to show a facing view of Sophie talking to the camera]
Some fluid flow problems are easy to solve like the single pendulum but a huge number of them cannot be solved like the double pendulum.
[Image changes to show a profile view of Sophie talking to the camera and then the image changes to show a facing view of Sophie talking to the camera]
But why do we care? Because chaotic fluid exists everywhere, in your coffee cup, in a river, in the atmosphere and yet we can’t predict what it does. If you want your bridge to stay up, or your plane to fly smoothly, then we need a better idea of how turbulence works.
[Image changes to show a profile view of Sophie talking to the camera]
And these problems are too hard to solve ourselves. We need to use other methods to try to understand what is happening.
[Image changes to show a facing view of Sophie talking to the camera]
We can use supercomputers to solve the equations that govern all fluid flow.
[Image changes to show the Navier-Stokes equations on a light blue screen]
All fluid motion is governed by a set of equations called the Navier-Stokes equations. They’re basically Newton’s Second Law as applied to the movement of a mass of fluid.
[Image changes to show Sophie looking at a whiteboard and the camera zooms out to show Sophie writing equations on the whiteboard and the camera zooms in on her hand, and then on the equation]
The Navier-Stokes equations are too complicated to solve by hand. In fact, the Clay Mathematics Institute in the US will pay you $1,000,000 USD if you can solve them.
[Image changes to show Sophie working on a laptop computer]
However we can use supercomputers to find extremely good approximations to solutions for us.
[Image changes to show a facing view of Sophie talking to the camera and then the image changes to show a profile view of Sophie talking to the camera]
Rather than being able to write down a solution that tells us what will happen at every point in space for all of time, it will compute the values of the velocity and pressure at the points in space for the times we tell it to.
[Image changes to show a facing view of Sophie talking to the camera]
An example of this is the flow over an aerofoil or a plane wing.
[Image changes to show a diagram of air flow over a plane wing and text appears: Laminar boundary layer over a ‘nice’ wing]
What we want is for the incoming air to flow smoothly around the wing and detach off at the back in a nicely controlled way. What happens in reality is that the layer of fluid can detach early, before it reaches the trailing edge of the wing.
[Image changes to show a digital image of turbulent air movement behind a plane wing]
When that happens, the detached layer of fluid gets buffeted around and forms vortices, which become turbulent, creating a turbulent wake behind the wing, which then causes drag.
[Image changes to show a facing view of Sophie talking to the camera and then the image changes to show a profile view of Sophie talking to the camera]
This is exactly what we do not want to see happening on the surface of an aeroplane. More drag means increased noise and energy dissipation, making planes less fuel efficient and compromising control.
[Image changes to show a facing view of Sophie talking to the camera]
Plane wings are a complicated geometry for a computer to deal with if you want to solve the Navier-Stokes equations in their full form.
[Image changes to show diagrams of three dimensional shapes mimicking air flow around those shapes and a scale can be seen in the centre of the screen and the image shows seconds counting up]
So I spin three-dimensional shapes that are easier to deal with mathematically in computational fluid to mimic the physical processes that happen on the surface of more complicated objects like an aerofoil.
[Image changes to show a facing view of Sophie talking to the camera]
Supercomputing is only growing as a key part of the scientific process.
[Images move through of various digital maps of the Earth and also various digital medical models]
A lot of the understandings we need to improve our world, like new technologies, medicines and industrial practices, will come from the simulations and models that research produces with supercomputers.
[Images move through of various simulation modelling maps and modelling tools for extreme weather events]
The way we track flooding, bushfires, drought, and extreme weather events is evolving right now as big earth-observation satellite data becomes plentiful and easily accessible.
[Image changes to show a facing view of Sophie talking to the camera]
Supercomputers are being upgraded all the time which is exciting for the future.
[Image changes to show a profile view of Sophie talking to the camera and then the image changes to show a facing view of Sophie talking to the camera]
For example the World’s First Market-Ready Diamond-based Quantum Accelerator will be installed at the Pawsey Supercomputing Centre.
[Image changes to show a profile view of Sophie talking to the camera]
This means it harnesses synthetic diamonds to build quantum accelerators.
[Image changes to show a facing view of Sophie talking to the camera and then the image changes to show a profile view of Sophie talking to the camera]
This will develop cutting-edge applications in machine learning, logistics, defence, aerospace, quantum finance and quantum research.
[Image changes to show a facing view of Sophie talking to the camera and then the image changes to show a profile view of Sophie talking to the camera]
In the supercomputing industry, scientists leading the research are equally as valuable as the people managing the systems, providing support, and building new tools for analysing scientific data.
[Image changes to show a facing view of Sophie talking to the camera]
To help explain how supercomputers work, we now invite you to pause this video to undertake a couple of exercises.
[Music plays and text appears on a blue screen: Take a pause and learn more about how supercomputers work]
[Image changes to show new text on a dark blue screen: A STEM career pathway]
[Music plays and image changes to show a facing view of Sophie talking to the camera]
So, I’ve always been interested in STEM because I always liked learning new things but I didn’t really know what I wanted to do when I grew up. So, when I went to university I decided that I really liked physics and that Space was cool. So, I thought I would become an astrophysicist. But then going to university I had to do all the maths I needed to do the physics and I realised it wasn’t actually the physics that I enjoyed. It was applying maths to physical problems. So, it was taking mathematics, applying it to the real world to solve those problems, and that’s Applied Mathematics. And so, I accidentally became an applied mathematician instead of an astrophysicist but it’s worked out really well.
[Image changes to show Sophie walking along the footpath outside the NCI building and then the image changes to show a close view of the building]
So, I really discovered my love for STEM as a six year old.
[Image changes to show a close view of Sophie walking past the NCI sign and then the image changes to show Sophie walking into the entrance of the Australian National University]
I loved sharks and I loved the ocean and so I decided that I wanted to be a marine biologist.
[Image changes to show a facing view of Sophie entering the university and the camera zooms in on Sophie as she walks towards the camera]
But then I worked out how scary sharks were and it just seemed a little bit too much and I ended up more at the physics and maths end of STEM. So, in Year 11 and 12 I did those STEM subjects, along with some Art subjects too because I really loved those.
[Images move through to show a profile view of Sophie at work, a close view of Sophie’s hands typing on a laptop, and then Sophie at work again]
So, at university I did a Science/Arts double degree so I could do astrophysics as well as picking up all of those other Arts subjects that I was still really interested in.
[Image changes to show a close view of Sophie taking a sip from a coffee cup and then the image changes to show Sophie talking to the camera]
And then it was when I was at university that’s when I found out I didn’t want to become an astrophysicist.
[Image changes to show a profile view of Sophie talking to the camera]
And in fact it was taking the maths and applying it to the astrophysical problems that I really enjoyed and then that became an all-encompassing thing.
[Image changes to show a close profile view of Sophie at work]
I realised I could take maths and I could apply it to any real world situation and I could solve that problem.
[Images move through of a close view of Sophie’s hands typing on the laptop keyboard, Sophie looking down, Sophie typing again, and then a rear view of Sophie working on a laptop]
After my Science/Arts degree I did Honours in Applied Maths and I looked at problems in neurophysiology. So, we were looking at mathematical models of neuron firing, so neurophysiology from a mathematical point of view and I had no idea I could do that. So, from then on I was an applied mathematician and I got into fluid mechanics.
[Image changes to show a facing view of Sophie talking to the camera, and then the image changes to show a profile and then facing view of Sophie talking to the camera]
And it means that now I can still study sharks, and I can study fluid flow around sharks, and why they move so fast, and why they’re designed the way that they are, or why nature designs them the way that they are without having to get up too close and personal to sharks like the six year old me was terrified of.
[Image changes to show a profile view of Sophie talking to the camera]
So, now as an applied mathematician I can work with everyone.
[Image changes to show a facing and then a rear view of Sophie walking along past the GADI supercomputer]
I can dip my toe into all the sciences that are out there. And what’s amazing about this is we don’t know what’s coming.
[Image changes to show a close view of Sophie walking down the data hall in the GADI supercomputer]
We don’t know what the problems are we need to solve in the future.
[Image changes to show a view of Sophie bending down and looking at something in the data hall and then walking towards the camera]
And because of my background in maths and my ability to use a supercomputer it means that I’ve made myself future proof and I’m really excited about the things that we’re going to tackle next.
[Image changes to show a facing and then profile view of Sophie talking to the camera]
I am so jealous of the future scientists of the world because there are so many different things out there.
[Image changes to show Sophie exiting one side of the data hall of the GADI super computer and then entering the other side of the data hall in the GADI super computer]
So, my advice to everyone would be try as many things as possible, keep your options open, embrace your passions, because what you don’t want is a career that you’re stuck in.
[Image changes to show Sophie leaning on the supercomputer and smiling at the camera]
You want something that you love to do every single day.
[Image changes to show a facing view of Sophie talking to the camera]
So, students think about how you could use super computers in your future career.
[Image changes to show a profile view of Sophie talking to the camera and then the image changes to show a facing view of Sophie talking to the camera]
This path can lead you into a range of exciting STEM jobs such as a research scientist, engineers, computer operators and technicians, visualisation programmers, and include jobs that haven’t even been invented yet.
[Image changes to show a profile view of Sophie talking to the camera and then the image changes to show a facing view of Sophie talking to the camera]
Remember mathematics is crucial to a career in STEM and working with supercomputers. Have we sparked a new interest in you and supercomputers?
[Image changes to show a profile view of Sophie talking to the camera]
There are plenty of places you can go to learn more about this topic and explore this career.
[Image changes to show a facing view of Sophie talking to the camera]
We encourage you to pave your own STEM path.
[Music plays and the image changes to show the CSIRO logo and text on a white screen: CSIRO, Australia’s National Science Agency]
[Image changes to show the NSW Government and the Science and Industry Endowment Fund logos and text appears: Generation STEM is managed by CSIRO and made possible by an endowment from the NSW Government to the Science and Industry Endowment Fund]
[Image changes to show the NCI Australia logo and text appears: Thank you to our content collaborators, NCI Vizlab, Professor Richard Sandberg, Professor Andy Hogg, Professor Todd Lane, Dr Claire Vincent, Dr Alejandro Di Luca, professor Jason Evans, Professor Hrvoje Tkalcic, Professor Ben Corry, Bureau of Meteorology, ARC Centre of excellence for Climate Extremes]