Name: The future of supercomputers? A quantum chip colder than outer space | Jerry Chow | TED Institute
Uploaded: 2020-03-20T01:41:45.000Z
Duration: 11 min 22 s
Description: Thousands of YouTube videos with English-Chinese subtitles! Now you can learn to understand native speakers, expand your vocabulary, and improve your pronunciation...

future progress in modern biology and medicine faces some astronomical hurdles.

But first, let's all sit back, relax and have a sip of coffee.

Now the active ingredient in coffee that keeps us chugging along keeps us awake.

Caffeine is a moderately sized molecule made up of a number of different Adams.

It's more complex than water, but simpler than, say, DNA or proteins.

Well, let's say we wanted to fully describe just this caffeine molecule, for example, to understand its internal energies, its structure and interactions to get her sense of how these atoms are working together to make caffeine what it is and how it stimulates us.

Well, just the complexity involved with this caffeine molecule is beyond the capability of being simulated by any computer today and in fact, beyond any computer ever possibly made with current technology.

Well, even you should take a computer chip, start packing in more and more transistors and make this computer really, really big and equal to the number of atoms that make up our planet.

Even you kept going and made this equal in to the number of atoms that make up our solar system, that wouldn't be enough, and it evens you kept on going, made this computer ship equal in size the number of atoms that make up our entire Milky Way galaxy.

You'd still fall short of just simulating that caffeine molecule.

So then, what does that say about more complicated and complex molecules in compounds?

Ones that are actually important for biological science and for battling diseases molecules like DNA or viruses, proteins and enzymes?

As great as our computers are today, there is a wall.

There's a fundamental limitation off our ability to simulate and understand the details structures interactions of these important and crucial, biologically relevant molecules in compounds simply because they're too big.

There's too many total number of atoms and electrons that make them up for our computers today to handle.

Now, the implications of this limitation are huge.

Right now, we're in a world where we always hear about ideas like person specific drugs or universal cures for aggressive cancers, or stopping viral epidemics or trying to figure out the origins behind life itself.

Now, current progress and all of these fields hinges upon trial and error.

And as great as some of the computation abilities of buying from attics and sequencing the genome have been for these applications, conventional computers just won't cut it.

We need processors beyond anything we've seen before.

We need a computing paradigm shift equivalent to taking us from the horse and buggy all the way out to warp drive on the Starship Enterprise.

This is a quantum processor, one of the cores of the quantum computing systems that my team is building, and this will drive the next generation of computing now, my whole entire life, I've wanted to understand how things work to make connections between physics and the real world, that we all know how important physics is for describing things around us, like how things fall, how the planets move or how we light up in power our homes.

But I never thought that something as esoteric and weird as quantum mechanics could be used to reveal how we as living beings, work and inside this very package are tiny circuits that obey the laws of quantum mechanics.

and have the potential to unlock real chemistry and Rio biology.

Now the reason for this connection is actually quite profound, and we have the throw it back to 1982.

One Nobel Prize winning physicist Richard Feynman issued forthis Revolutionary Challenge.

He basically said, You know, if we're gonna try and simulate these naturally occurring molecules, well, it be silly to do that with these computers that just aren't built for that challenge.

Instead, we should use a different set of objects, which we have control over and which follow those same exact laws of quantum mechanics.

There's that connection and I've been searching for, and that's how I want to apply physics to our real world Now.

It's interesting at the end of this quote that Fineman says it doesn't look so easy.

And that's certainly proven true over the last 20 plus years because you see computing with quantum mechanics is dealing with information that is exquisitely delicate and fragile.

With a regular computers, we process information as bits as zeros and ones.

It's just like setting a bunch of switches on and off, and we set them they say where we want.

But in quantum computing, information isn't just binary.

It could be one or could be in a superposition of zero and one the same time.

Now, I know that sounds weird, but welcome to the wacky world of quantum computing so that we have all these additional options for the information you zero could be.

Want to be anywhere in between on the surface of this sphere.

And we refer to this unit of information as a quantum bit or Cubitt.

The challenges that cubits are fickle little beasts way trick you to simply set and forget the way we do with those switches in our regular computers.

Trying to preserve one of these funky superposition states in a Cuban is a bit like trying to balance an egg at the end of a needle.

But any little disturbance from noise from heat from vibrations and you certainly got yourself a sunny side up.

So what this all means is that there are three critical challenges for making real physical quantum computing systems.

First, we actually have to have cubits that we can set the positions for second once we've got them set, preserving them, keeping them right where we want.

And third, bringing together many of these cubits all interconnected so that we can start to process and calculate some of those difficult molecules.

And in fine Wednesday, this holy trinity of controllability preservation and scalability was simply impossible.

And then I am standing in front of you here today because over these years, the impossible is finally starting to become possible.

Every year born, more quantum competing teams throughout the world are making gigantic strides towards five mins.

Dream, for example, in my lab, we are building riel physical cubits.

Well, actually, they're superconducting Cubans, which just means that they're made out of materials that need to be really, really cold in order work The great thing about the excuses that they run on simple electrical signals giving us controllability.

And second, by being really, really cold.

But now when I say cold, I mean really, really cold.

I mean, like minus 459 degrees Fahrenheit or minus 2 73 for the Celsius speaking world.

But that's colder than outer space itself, and that we have to make our cubits precisely this cold to protect and preserve those delicate superposition states that give quantum computing its power.

Now, 15 years ago was the first demonstration of this type of technology.

That time the world was stunned and it held a quantum information for just a nanosecond.

Now, today we're building cubits that lasts for over 100,000 times longer.

We test them in these refrigeration systems that can essentially run for as long as we want.

Those white cylinders are fridges which hold the Cubans that we build.

So now, with all this progress in controllability and preservation, return our attention to scale ability.

Early this year we demonstrated this four cubit processor and he's actually the one that's in my pocket here.

But the great thing about this four Cuban device is that the four cubits, which are the four squares of the middle, are arranged in a way that is scalable towards even larger numbers of Cuba's.

This is a building block, and we've already taken the first that beyond and are looking at this age of a device right now as I speak.

So they're already at this eight cubic processor level, we can start to do some simple chemistry, like looking at the bonding properties of a molecule like hydrogen h two.

But then, what about those more complicated molecules that we talked about, like caffeine or proteins and things like that?

But this is an important milestone to mark because with a little more elbow grease, heck of a lot more caffeine to clear, just a few more engineering hurdles soon will put together networks of 20 to 100 cubits in a processor.

And when that happens, then we can take on some of these problems that surpassed the capabilities of our current computers.

The focus shifts from Can we build this darn thing to How do we use it?

And to me, this is the most exciting part about quantum computing.

If is where we can take the discourse beyond physics and engage with fields like chemistry, like biology, with groups of people working on the most challenging and difficult problems in human health in the natural environment.

And just as conventional computers had revolutionized our world before us.

 So we'll quantum computers when we bring together this powerful mind share and reignite our collective scientific imagination. 