TED Talks · Level C1 Mixing - Part 2

Have we reached the limit of computer power

In the Netherlands, there's an ambitious company that builds one of the most advanced and expensive tools in the world: a single unit costs hundreds of…

Level: C1 · 74 sentences

Transcript

In the Netherlands, there's an ambitious company that builds one of the most advanced and expensive tools in the world:
a single unit costs hundreds of millions of dollars.
And when companies buy one, they also need 250 engineers
to install the 165-ton device in a process that typically takes half a year.
But despite this steep cost in time and money,
many microchip makers desperately want one of these machines.
The hundred-million-dollar question is: why?
The answer has to do with something called Moore's Law.
First coined by Intel co-founder Gordon Moore,
this law states that every 1 to 2 years,
the number of transistors that can fit on a given size computer chip will double.
And by extension, the rough number of calculations that chip can do per second will also double.
Now, this law isn't a physical law like gravity.
It's just a trend Moore observed during the early 1960s.
But chipmakers turned that trend into a goal,
and in turn, consumers learn to expect computing progress to continue
at this exponentially fast pace.
And the amazing thing is, for six decades, it pretty much has.
Thanks to Moore's Law, chips have gotten smaller, faster, more efficient, and cheaper.
But today, there are four key problems that trip up this trend,
potentially ending Moore's Law
and fundamentally changing how we make progress in computing.
The first is transistor size.
Transistors are basically on/off switches,
and these building blocks of digital computing have been shrinking since the 1960s.
But recently they've gotten so small,
quantum physics has begun to interfere with their functions.
When a transistor's switch, or gate, is less than 20 nanometers,
electrons will tunnel along it continuously,
turning a crisp on/off switch into a hazy dimmer.
The second problem is heat.
As chipmakers make components smaller and more complex,
the copper lines that run between them need to be thinner and longer.
This increases their electrical resistance and generates high heat
that impairs chip performance and can't be easily dissipated.
Today's chips can already run hot enough to cook an egg,
and temperatures are only predicted to increase without new innovations.
While both these issues represent limits in the fundamental physics of chipmaking,
researchers haven't stopped trying to solve them.
Unfortunately, their solutions often exacerbate the third major problem:
chipmaking's environmental impact.
For example, swapping copper lines for ruthenium
could help pack transistors more tightly and keep chips smaller,
but that metal is far scarcer than copper
and would require new mining infrastructure.
Similarly, the technology currently used to make today's smallest transistors
requires huge amounts of energy and chemicals called perfluoroalkyl and polyfluoroalkyl
substances, which can take thousands of years to break down in the environment.
Managing these first three problems contributes to the final issue: cost.
To keep achieving Moore's Law,
chipmakers have to make individual chip components smaller.
And this is where that costly 400 million dollar machine comes in.
This marvel of chip-making science shoots a stream of tin droplets into a vacuum chamber
before blasting them with a high-energy laser that vaporizes the tin to create plasma.
In turn, the plasma emits a 13.5 nanometer wavelength of ultraviolet light
that can be used to produce incredibly small transistors.
This remarkable feat of engineering has helped chipmakers keep up with Moore's Law.
But as chips keep getting denser,
intricate manufacturing plants keep getting more expensive.
This trend has been so consistent, it's actually earned the nickname Moore's Second Law.
Obviously, all these trajectories are unsustainable.
Manufacturing plants can't keep increasing in price,
our ecosystems can't endure endless mining and pollution,
and the laws of physics are unlikely to change anytime soon.
Fortunately, Moore's Law is flexible,
and there's no reason we can't introduce new goals
to keep making computing progress responsibly.
Perhaps we could introduce a new Sustainability Law?
Smaller transistors already use less material and produce less e-waste,
and advancements in electronic-photonic integration are allowing chips to use less energy
and generate less heat.
So perhaps chips should be made twice as sustainable every several years?
Whatever the answer is, we make the laws,
so the future is up to us.