Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
Hey, please take a second and leave us a review
on Apple Podcasts, Spotify, or wherever you listen to the podcast.
Thanks a lot, Hey, Happy New year. In case you
didn't know, twenty twenty five was the Year of Quantum,
the year where we celebrated the one hundredth anniversary of
the development the theory of quantum mechanics. So to end
(00:20):
the year, we're going to re air our episode that
answers the question what is a quantum computer. It's a
good episode because it's a nice introduction to quantum ideas
and you'll hear what it's like to go inside an
actual quantum computer lab. We'll be back next week with
a new episode about the science of imagination, What is
it and how does it work in your brain? So
(00:43):
be sure to come back next week for that, But
until then, please enjoy this field trip into the quantum world. Hey,
welcome to Sigence Stuff, a production of iHeartRadio. My name
is hoorheitch Ham and to the end of the program, we
are talking about a technology that made potentially impact the
life of every single human on Earth and might change
(01:05):
how we protect data and come up with passwords. It
might help us make new and exciting materials, and it
might render cryptocurrencies like Bitcoin and dotgecoin totally useless. I'm
talking about quantum computers. What are they, how do they work?
And most exciting is that we're going to get to
visit one of them and actually hear it in action.
(01:26):
So power up your curiosity, log in, and let's answer
the question how do quantum computers work? Hey? Everyone? Okay,
So when I started this episode, I was both terrified
and excited. Terrified because explaining anything with the word quantum
is really hard, but excited because I had heard that
(01:49):
a friend of mine was making quantum computers just ten
minutes from my house and this was a great excuse
for me to go take a look at them. So
we're going to go see these quantum computers in person
at the end of the episode, but before that, I
wanted to make sure that I understood what they were,
how they work, and also what they're potentially going to
be used for. So this episode is split into three parts.
(02:12):
What is a quantum computer and how does it work?
What are quantum computers for? And then we're going to
go see the quantum computers. We're going to talk about
how they're made and why they're so hard to get
them to work. Our guide through all of this is
going to be my friend who's making the quantum computers,
Professor Oscar Pater. He's a professor of physics and applied
physics at Caltech and he's the head of quantum hardware
(02:35):
for Amazon. He does research on nanophotonics, quantum optics, and
of course, quantum computers. Here's my visit to Oscar Painter's lab. A. Hey,
os Care, how.
Speaker 2 (02:52):
Are you good to see you after so many years.
Speaker 1 (02:54):
Yeah, it's been a while. Huh yeah, well thanks so
much for talking with me.
Speaker 2 (02:57):
Yeah, it's been a while. Happy to try to fill
you in on some of the things we've been doing
in the areas of quantic computing.
Speaker 1 (03:04):
Okay, so the first thing I wanted to talk to
him about was just what does the word quantum mean?
Because I feel like we're going to need that to
understand what a quantum computer is.
Speaker 2 (03:14):
Now.
Speaker 1 (03:15):
The word quantum is the word we used to describe
how things behave at the level of atoms and the
tiny little particles that make up the atoms. So in
our everyday lives, we're used to things being solid and
us being able to hold them. Like, for example, if
you take a piece of wood or a ball. But
if you take that piece of wood or ball and
you chop it up, and you keep chopping it up,
(03:36):
you get down to atoms, and then you'll notice that
those atoms don't behave in the same way that a
piece of wood or a ball do. Here's how Oscar
explains it.
Speaker 2 (03:45):
It turns out that down to the microscopic scale, so
not our everyday scale of things, the laws of physics
that dominate in that regime is quantum mechanics, and quantum
mechanics is our theory that has some strange attributes that
we don't experience every day. Exact ample, it postulates that
things can be in superposition, so you can have objects
being in sort of what we think of as two
(04:07):
distinct realities at the same time. Imagine having a particle
in one position and another position simultaneously. That seems very
odd to us, but in quantum mechanics it's very natural.
Speaker 1 (04:18):
Like, for example, I grabbed this piece of wood in
front of me, and it's a piece of wood. It's
not two things at the same time, right, It's in
one location.
Speaker 2 (04:25):
Right, it's sitting there firmly right in front of you.
Speaker 1 (04:27):
Right, if I had an atom in front of me
or an electron, it wouldn't exactly.
Speaker 2 (04:32):
You would find that if you repeated the measurement or
finding its position multiple times, you might find that, Oh,
I get this weird outcome that sometimes I measure it here,
sometimes I measure it there. And that's because it's actually
in many places at once, all right, And that's fundamental
to the description of quantum mechanics. The way I like
to think about quantum mechanics is really as waves and amplitudes.
(04:53):
So think about you're at a pond and you throw
a rock in a pond, and you see this ripple
of the rock. Right, That's how I think about, Like
the rocks are sort of the particles, and these wave
phenomena are sort of the actual physical quantum mechanical description
of that particle.
Speaker 1 (05:08):
Like the particle, the thing, the atom, or the electron.
It's not the rock you throw into the pond, no,
but it's actually the ripple of the ripple. Yeah, that's right,
it's this wave.
Speaker 2 (05:17):
So I may have started with something that was very local,
like that rock, but then it becomes very quickly it
sort of propagates out and is actually better described as
this wave on the pond, because.
Speaker 1 (05:26):
Like a ripple and a wave in a pond like that,
it's kind of in a lot of places at the
same exactly.
Speaker 2 (05:31):
That's right. And then the interference is important to understand.
If I throw two rocks in the pond, then I
see the sort of interference of the ripple patterns coming
from each rock that flashed in the pond.
Speaker 1 (05:41):
Right, Like each ripple starts at simple, but then they
start to mix together and form this complex pattern on
the surface of the pod.
Speaker 2 (05:50):
Exactly, like how do they evolve in time?
Speaker 1 (05:54):
Okay, so when you get down to the level of atoms,
things behave really strangely. Scientists think of things at that
level not as little tiny balls, but as waves or
ripples of energy, like the ripples in a pond. Now
you may think, wait a minute, if things are kind
of wavy and strange at the level of atoms, why
isn't it that way when you get to big stuff
(06:15):
like a piece of wood or a ball. And the
answer is that they are that way. There's just a
lot of atoms in a piece of wood, and from
a distance, it gives you the impression that it's solid.
It's sort of like how some clouds from Afar they
might look solid, once you get up close to them,
they're actually kind of fuzzy and wispy, and all the
(06:36):
water droplets are moving around. So that's quantum. Now, a
quantum computer is what happens when you make a regular computer,
but you make the circuits out of individual atoms or
particles like electrons.
Speaker 2 (06:51):
Classical computers are formed from things that are very very
classical in nature.
Speaker 1 (06:54):
And they operate kind of on hard switches.
Speaker 2 (06:57):
Yeah, like, yeah, that's right. The transistors on your phone
what we call these types of elements. The transistors are
used to store information or perform calculations, and the transistors
are really set by a bunch of electrons in part
of the circuit. And usually you're talking about quite a
few electrons.
Speaker 1 (07:12):
Because regular transistors are huge. They're bigger than an atom.
Speaker 2 (07:16):
Yes, exactly, that's physically what's going on in your phone.
And what I'm telling you is that the way to
think about it is in the quantum case, I just
have one electron.
Speaker 1 (07:24):
Like, the circuits are made out of individual electrons.
Speaker 2 (07:27):
Yeah, atoms is exactly what you're doing or a quantum
particle doesn't have to be electrons, can be other particles.
That was the sort of very early idea from people
like FIM and others back in the nineteen eighties is
if you're going to do this, and you better make
them out of quantum mechanical objects to begin with.
Speaker 1 (07:43):
Okay, So if you make a computer where the circuits
are made of individual quantum objects like atoms or electrons,
then you get a quantum computer. And what that does
is that it makes their calculations also quantum mechanical. And
this is where the concept of a bit comes in.
It's like a regular bit in your computer, and a
(08:04):
bit is like a one or zero, but a cubit
is a quantum mechanical one or zero.
Speaker 2 (08:11):
Classic computers are formed from digital bits and they go
between one and zero. A quantum computer doesn't have these
hard zero one states. It has every possibility in between.
So imagine if we have these two states zero in one.
I told you that a quantum system can be in
two different states at once, right, So it can be
in zero and one at the same time, and I
(08:32):
can have a different weight of zero or one at
the same time. It could be ten percent zero ninety
percent one.
Speaker 1 (08:38):
It's like a shade of gray.
Speaker 2 (08:39):
Yeah, And so you have all those possibilities in between.
It can be zero or one or anything in between.
Speaker 1 (08:46):
It can be black, white, dark, gray, light gray exactly.
Speaker 2 (08:49):
So it has all those shades in between. You can
take zero with some fraction and add it to one
with any other fraction. You can have any combination of that.
Speaker 1 (08:57):
So an a regular computer, if you multiply two bits together,
it's like you're multiplying two fixed numbers together, like three
times four. But in a quantum computer, when you multiply
two cubits together, it's like you're multiplying two things that
can be lots of numbers at the same time. So,
for example, it's like you're multiplying every number from zero
(09:18):
to one hundred times every number from zero to a thousand,
all at the same time in one operation. That's what
makes quantum computers unique. They take this weirdness of the
quantum world and it lets you do math with it. Now,
actually it's not doing all of those multiplications or calculations
at the same time. It's more like how Oscar described
(09:39):
it earlier. If you drop two rocks in a pond,
you see the two ripples spread out and mix together
to form a complex ripple pattern. That's more of the
picture of what a quantum computer does. It doesn't do
calculations with hard numbers. It does calculations with the ripples
and patterns of quantum numbers. Of course, my next question
(10:00):
for Oscar was what is that good for? Why would
you want to do math? This way? When we come back,
I'm going to ask Oscar what quantum computers are for,
and then at the end we're going to go check
out the ones he's built. You're listening to science stuff.
(10:24):
Welcome back. Okay, to recap, we learned that a quantum
computer is a regular computer whose circuits are made with
individual atoms or small particles like electrons, and by doing
that you can do quantum calculations. That is, you can
do math, but with numbers that are actually lots of
different numbers at the same time. So now the question
(10:44):
is why would you want to do that? What are
quantum computers? Four? Here's more of my conversation with quantum
physicists Oscar Painter. Let's say it's a few years into
the future and we have quantum computers, yes, in our phones,
like I have one in my pocket, what can I
do with it? And how is my life different?
Speaker 2 (11:04):
I think that's a very unlikely scenario. Okay, I think
that's the wrong way to think about how quantic computers
might change our lives, at least as far as I
can project into the future. I think the best way
to think about a quantic computer as we envision it
right now is that it will be more like a supercomputer.
(11:25):
So a supercomputer is just a very large computer that
can perform calculations beyond what our desktop, our personal computers
can do. And these are usually very large, almost building
scale computers and computer clusters that have many, many different
processing units that are all integrated together, and through that
scale you can perform a huge number of computations per
(11:45):
second and therefore compute some of the hardest problems that
are out there. A lot of them are used for
chemistry problems. They're used to study particle physics, so fundamental physics,
trying to understand models of quantum particles that are beyond
the current standard model. There used to compute the properties
of materials, climate modeling, and things like that.
Speaker 1 (12:04):
So usually science and tech, yeahs, and.
Speaker 2 (12:07):
There's always this competition between different nations who has the
fastest or the biggest supercomputer.
Speaker 1 (12:12):
I see, So you envision quantum computers will be sort
of like a specialized version of computer.
Speaker 2 (12:18):
It's going to be some very special type of supercomputer
that can solve specific problems that quantic computers will be
very effective at that we can't do today on classic computers,
no matter how much we scale them up. And the
key is, it's not just a faster supercomputer. It performs
calculations in a fundamentally different way, and therefore it can
tackle problems that are possibly outside of the reach of
(12:39):
these conventional classical supercomputers. What do you mean, out of
the reach meaning that no matter how fast they get
or how big they get, they'll never be able to
compute some.
Speaker 1 (12:48):
Of these problems, never or take an infinite Yeah, exactly.
Speaker 2 (12:52):
It's just the scaling is so bad for these problems.
You would take way too long and require way too
large on machine. So no matter how hard we work
on our current computing technology, it has limits, and it's
known and you can prove it for certain problems. And
quantic computers when they looked at theoretically these same problems,
(13:13):
they realize that the same restrictions or limitations for quantum
computers are not there. There's examples where we believe and
strongly believe that certain mathematical problems that are important are
really really hard to perform and can't be solved using
classical means, no matter how much we improve the technology.
Speaker 1 (13:33):
No matter if I have a building full of supercomputers,
yeah exactly, you'll never be able to, say, just fill
the world with them.
Speaker 2 (13:38):
You still won't be able to do it. Yet a
quantic computer can solve it pretty efficiently.
Speaker 1 (13:43):
Well, step me through some of these problems.
Speaker 2 (13:45):
Like so the example that everyone points to, and it
is pretty amazing that people found this, But there's this
mathematical problem and it just happens to be very applicable
to our safety or security of our data. So it
turns out that most of the security of all of
the data you hold, all the data that banks or
various institutions around the world want to be safe and protected,
(14:08):
they typically encrypt it. And those encryption techniques that have
been used were what are called RSA encryption, where you
want to take a large number and understand what its
prime factors are.
Speaker 1 (14:20):
Okay, so the first big thing that quantum computers can
be useful. The one that got people really excited about
them in the nineties is in breaking password encryption. So
whenever you enter your password on a website or when
you download your bank statement, that information is encrypted or
scrambled so that if anyone happens to catch that information,
(14:41):
they can't tell what it says. And the whole scheme
is based on the idea that if I gave you
a really large number, it's really hard to find what
its prime factors are. Here's how Oscar explains it, but
just to give you a quick heads up, a prime
number is a number that can't be divided except by
itself or by one. So, for example, thirteen is a
(15:04):
prime number because you can't fight thirteen by anything except
thirteen and one. And the same goes for seventeen, nineteen,
twenty three and so on. Anyways, here's Oscar explaining it.
Speaker 2 (15:17):
So I give you a number, and I say, tell
me what the prime factors are, and you have to
break it down to its prime factors. So you know,
a simple one is, you know, like two, it's just
one times two, one and two those are the two
prime factors, right, But it gets harder as these numbers
get bigger.
Speaker 1 (15:31):
If I tell you one million, three hundred forty three.
Speaker 2 (15:34):
Yeah, thirteen, very hard to actually answer what those what
the prime factors are. But if I give you the
prime factors, you can multiply them together and very quickly
get the answer to what that larger number is.
Speaker 1 (15:46):
Right.
Speaker 2 (15:47):
And so if you know the prime factors, I can
give you what they multiply to. But if you give
me the number that they multiply to without then I
have a very hard time finding out what the prime factors.
Speaker 1 (15:56):
Are because you'd have to get you kind of have
to guess.
Speaker 2 (15:59):
Well, you know, there's mathematical techniques to try to find these,
but they're very inefficient. And so it turns out that
most of the security of the way we encrypt information
is based upon that asymmetry and how hard the problem is.
Speaker 1 (16:13):
So now, let's say somebody has a quantum computer.
Speaker 2 (16:15):
Right, then they can find those prime factors and they
can now decrypt all that information.
Speaker 1 (16:22):
They can just grab it from the air, yeah and
be like, oh, I.
Speaker 2 (16:25):
Know, yeah, I can find the prime factors and then
I can use that to decrypt the information.
Speaker 1 (16:32):
That would be easy for a quantum computer you just
press a button and will tell you, Oh, this is
an oscar or his secret decoder.
Speaker 2 (16:38):
Yeah exactly, so that would you know. That obviously concerned
a lot of people when that algorithm was developed.
Speaker 1 (16:48):
Okay, this gets a little bit heavy into encryption and
quantum algorithms, but the main point is that most of
the security of our passwords and our sensitive information, and
also the encryption of things like bitcoin and all those cryptocurrencies,
they all depend on this one math problem which is
really hard for regular computers, even supercomputers to solve. And
(17:10):
that is a problem of finding the two prime numbers
that multiply to get a really large number. But then
in nineteen ninety five, a computer scientist named Peter Shore
publish the paper titled Polynomial time Algorithms for prime factorization
of discrete logarithms on a quantum computer, which essentially showed
that if you have a quantum computer, you can solve
(17:30):
this problem in a short amount of time. And this
is probably the main reason that people have been rushing
to make quantum computers since then, because imagine if everyone
in the world, people, companies, countries are all protecting their
secrets using the same trick but you had a special
quantum computer that could break that trick, you could rule
the world. Now, the details of how Peter Shore's algorithm
(17:53):
works are a little complicated to explain here, but the
essence of it is that you're using the ripples on
au pawn nature of quantum numbers on a quantum computer
to basically try out every possible combination for how to
break your secret encryption, and you use some clever math
tricks so that these ripples combine and mix together until
(18:13):
the right answer pops out. So that is the main
reason that people are excited about quantum computers. But there
are other reasons and other possible applications, So here's Oscar
telling me about them.
Speaker 2 (18:27):
Another example is maybe more natural to think about, and
this is where quantum computers were first proposed. It to
be interesting or useful, and that is the simulation of
nature itself. Nature as we know it is not classical.
If you peel the layers of the onion enough and
you get down to the core, right to the atomic scale,
it turns out that the laws of physics that dominate
(18:49):
is quantum mechanics. Okay, like the actual mathematics of that
When you describe it. When you have many particles, it
quickly becomes something that you can't simulate with a classical computer.
So all those interference of all the particles and keeping
track of all of that. A classical computer, if you
try to simulate that, you quickly run out of steam
and it becomes an exponentially hard problem. And so you know,
(19:12):
a classical computer is just ill suited to doing that.
But a quantum computer that's made out of the say
those sort of particles that can do with that interference naturally,
you know, has a natural advantage in terms of using
it to simulate the natural world at its quantum mechanical core.
Speaker 1 (19:25):
Why would I want to do that?
Speaker 2 (19:26):
Yeah, so that's the questions like, okay, so that's great,
But why would I want to do that other than
maybe I want to understand physics better. Well, this idea
that I want to understand how material behaves is a
very good example. If I'm building an electrical circuit, or
I'm building a new battery, or I'm building a different
energy process inside of a material or energy storage device,
A lot of times that depends on what the electrons
(19:47):
are doing. If I want to understand is there something
unique when I describe them quantum mechanically. Maybe there's special
properties I'm just totally blind to. So if I wanted
to make a better superconnecting material, something that can carry
a electricity with no resistance, right, maybe we can have
magnetically livitated trains. Maybe you can have you know, really
efficient electrical circuits that don't dissipate any energy. All of
(20:10):
these things. Then I would have to use a quantic
computer to model that behavior.
Speaker 1 (20:14):
And you said there's some maybe potential applications in chemistry
and biology.
Speaker 2 (20:17):
Yeah, you know, if I think about what is going
on when I have a chemical reaction, usually it comes
down to the electrons, and I need to understand what
they're doing in order to understand, you know, whether this
chemical reaction is going to be efficient or not, or
if I want to describe it with chemical accuracy so
I can use it to, you know, do some sort
of industrial chemical process. The biological application, it's like, if
(20:38):
I want to know how molecules are biologically relevant molecules
lined together, then potentially I need to know more information
about the electronic behavior in these molecules. If I wanted
to do that without having approximation or with much higher accuracy.
Then a quantic computer would be potentially more capable.
Speaker 1 (20:54):
There we might be able to predict better how a
vaccine will work, or whether a certain chemical into this
in your body will.
Speaker 2 (21:01):
Right now, we don't have that sort of level of specificity.
I mean, we'd love too. People are proposing techniques, but
that's the right idea, by the devil's in the details.
And you know, you have people saying, well, look, you know,
I think even today there's I won't call them skeptics,
but there's a lot of people that are saying, well,
I can keep improving my classical algorithms, and whether you
can really gain advantage from the quantum simulations is as
(21:24):
a practical question. And maybe we don't have as clear
an example or as clear a win when it comes
to how quantum computers will will do better or be
more efficient, or be able to do the calculations fast
or even do them ones that the classical computers can't do.
But I think there's definitely something there. It's just that
we still have to work on the quantum algorithms. It's
not as clear cut I would say.
Speaker 1 (21:45):
So those are the two main applications or uses for
quantum computers. One is in breaking encryption using a special
algorithm called face estimation that only works in a quantum computer,
and the other is to simulate nature, because nature is,
after all, quantum at its core, and so scientists think
that quantum computers will let us better simulate how atoms
(22:08):
and electrons interact, so that we can design better materials,
better semiconductors, and maybe better medicines. Now, I said so far,
because this is all still very new, and there might
be other classes of problems like the encryption problem, where
quantum computers are just fundamentally and exponentially better at solving,
(22:29):
but nobody knows for sure. Of course, it's all hinges
on whether or not we can actually make quantum computers
at the level that they would actually be useful and
most important, reliable. So now we're going to go actually
see these quantum computers that Oscar is building, and he's
going to tell us why they're hard to make and
why they're so prone to making errors. But first let's
(22:51):
take out quick break. You're listening to science stuff and
we're back. Well, I heard you have a quantum computer
in your basement. Well, not in my.
Speaker 2 (23:07):
Basement, but in my my laboratory. Yeah here, can we
go see it? We can?
Speaker 1 (23:12):
Okay, yeah, let's get see it.
Speaker 2 (23:13):
Okay, you want to do that now?
Speaker 1 (23:14):
Okay, so where are we going?
Speaker 2 (23:18):
Just the next door. We don't actually even have to
go down into the basement, into the basement, No, we
can the basement.
Speaker 1 (23:25):
Sounding is more yeahd scientists exactly.
Speaker 2 (23:28):
So let's so all these labs have different variants of
quantic computers that we're testing. Multiple quantum computers here, yeah, yeah,
not just one. So there's small scale quantum computers, but
the largest ones are you know, ones at Amazon or
Google or IBM or you know some of the other
(23:48):
startup companies. These get to be maybe a factor of
ten times larger than the ones I'll show you. Okay, okay,
so this gives you an idea all of these control electronics,
right is to use to an about twenty of these
quantum bits.
Speaker 1 (24:02):
There's twenty quantum twenty particle made up of twenty quantum
particles correct.
Speaker 2 (24:09):
Right, which we are manipulating as quantum bits, and that
circuit lies down inside of this special refrigerator.
Speaker 1 (24:16):
Okay, So if you've ever seen or if you google
a picture of a quantum computer, most likely what you
see is something that looks like an upside down metal
wedding cake with circular tears or platforms that get smaller
and smaller as they hang down from the ceiling. That
is basically a super intense refrigerator. The whole purpose of
(24:39):
it is to get the tip of that upside down
cake really really, really cold.
Speaker 2 (24:46):
And this refrigerator is under vacuum, under high vacuum. It's
a temperature which is about ten million degrees above apsoute zero.
Speaker 1 (24:53):
Ten million degrees.
Speaker 2 (24:54):
So to give you an idea, so if I go
to the deepest part of space, it's a few degrees calvin,
a few degrees about that food and zero the coldest
darkest parts of outer space or that universe. But this
thing's about thirty times colder than that.
Speaker 1 (25:08):
Even WHOA, So would you say that some of the
coldest places in the whole universe?
Speaker 2 (25:15):
I mean no, I mean you can get and there's
people that do this for a living that make really
cold things. But this is among the very very coldest things. Okay, yeah,
but this is extremely cold.
Speaker 1 (25:25):
What does it need to be cold?
Speaker 2 (25:26):
Because even the lights, even if we turned all the
lights off, even just the fact that the room's hot.
It's room temperature, but it radiates radiation, and that radiation
would completely destroy the information in the corner.
Speaker 1 (25:39):
Bit I see. So we have to get it really dark.
Speaker 2 (25:41):
We have to make sure that there's not any of
this thermal energy that's making it into the circuit, otherwise
it'll destroy the manipulation of those count particles. And so
it has to be as isolated as we can from
the environment. We would ideally seal it off from everything,
so it would be like zero temperature and there would
be nothing coming in other than what we want to
send to it to control it. And then you can
see there's all of these cables.
Speaker 1 (26:03):
Each of these feeds.
Speaker 2 (26:04):
Into a microwave cable that can be used to control
individual quantum bits or quantum particles on the circuit.
Speaker 1 (26:13):
So what I'm looking at is a room full of
electronics and cables, and in the center is a massive
structure with two suspended eye beams, and hanging from those
beams is the upside down wedding cake I mentioned before,
which in this case is sealed inside a really thick
metal cylinder, and inside that cylinder, at the very tip
(26:35):
of the wedding cake cool to almost the coldest anything
can be in the whole universe. Is a little chip
with a quantum computer. Well, what's in there? So describeting
what's inside the core of it? Is it like a
little chip? Yeah?
Speaker 2 (26:49):
Like we shit, Okay, it's what's called a superconnecting quantum circuit.
So it uses little metal traces on a silicon wafer
that we pattern on the surface, and when you get
them cold enough, they become superconnecting, which means they can
carry electrical currents without any energy dissipation. Okay, And it
turns out that you can form these sort of quantum
particles like these atoms where the current is circulating in
(27:12):
a clockwise way inside of a little tiny ring, or
it's circulating counterclockwise, and the clockwise could be zero when
the counterclockwise could be one, And you can get in
any superposition of these two circulation patterns, and I can
use that, and I can manipulate what the superposition is,
and I can have an interact with other circulating currents
to these little circuits things in our circuit. There are
(27:33):
a few hundred microns in size, so they might be
a few times the human hair diameter, so they're pretty
big relative to conventional transistors. It's made out of many atoms,
but it behaves like a single atom. Okay, yeah, theod
way to think about it. So there's like a little
array of these things, a little array of these things
on the surface of a microchip, and then each of
them we can control the current flow. So what are
(27:55):
called single cubic gates. We bring them together and then
let them interact them bring them apart. So I need
to build to manipulate the single particle, put it in
any sort of superposition I want. And then you have
to read out the state of these cubits too. You
have to know after I do my computation, are you
in state zero or state one? All right, I have
to ask that question for all my cubits and that
will give me the answer I see hearing.
Speaker 1 (28:15):
So this is called the pulse tube cooler.
Speaker 2 (28:18):
Some oder sounds like if you were ever a kid
growing up in the eighties and you watch Battlestar Galactica,
there were the cylons and Battlestar Galactica.
Speaker 1 (28:24):
They walk and they would.
Speaker 2 (28:26):
Have this flashing light and they'd have this sort of
sound coming from them. This is a similar sort of sound.
This is a pulse tube cooler and is shooting a
slug of helium gas onto a cold plate, and then
in doing so, when it expands, it can cause cooling.
It's analogous to what you do with the regular refrigerator.
That's the first stage of cooling, though that only gets
you down to maybe a tenth of the temperature of
(28:46):
the room. And then if I want to go even
cooler down and by another factor of ten or one hundred,
then you have to use a recirculating gas. In this case,
it's a dilution fridge that takes mixtures of isotopes of
helium helium three and helium four, and when they mix
there's an entapley of reaction and that's what gets you
down to this lowest temperatures I mentioned.
Speaker 1 (29:05):
So it's several stages. Something like take a fridge, put
it inside of another fridge.
Speaker 2 (29:10):
There's actually like sort of three or four stages.
Speaker 1 (29:12):
If you're inside of a fridge, you have because of
the fridge, the.
Speaker 2 (29:15):
One fridge is too hot for the other fridge, so
we have to isolate them, and then we have to
do that for every successive stage.
Speaker 1 (29:22):
It's like if I take my freezer and I put
it inside of like a restaurant freezer. Just be colder. Yeah, okay, exactly,
I keep doing that.
Speaker 2 (29:29):
I add, you know another.
Speaker 1 (29:30):
Way, the fridge inside of my fridge have a restaurant freezer.
Speaker 2 (29:33):
And I keep you know, each of them has the
ability to get colder and colder.
Speaker 1 (29:36):
Yeah, so you have to do it in stages.
Speaker 2 (29:37):
Otherwise, if you try to do a direct shot, it's
too much of a thermal load on the system.
Speaker 1 (29:41):
I see. So that's a quantum computer in action. Most
of what you see when you look at a picture
of a quantum computer is all the machinery needed to
keep the actual circuit in the near perfect vacuum and
as cold as possible, And all of that is to
completely isolate the quantum computer from the outside world. We'll
(30:01):
get to why you need to do that with a
quantum computer, but first I was curious how much a
quantum computer like this costs. Here's what oscars it. Well,
there is definitely much bigger than my phone. Yes, exactly.
Speaker 2 (30:15):
That's why I was saying, you're probably not going to
carry one of these things around.
Speaker 1 (30:17):
How much is this something like this? If I wanted
to build one in my.
Speaker 2 (30:20):
Garage Okay, Well, you know there's always a big difference
between science money and money that you know, when you're
talking about conservative products that have large volumes. I remember
the first time we purchased a big piece of equipment
from my lab, when I was the first to saculty member.
It was a bolt the same It was about smaller
than this thing, so smaller than a few cubic feet,
but it was more expensive than my.
Speaker 1 (30:40):
House when I bought it. So there's a big difference.
Speaker 2 (30:43):
But so just keep that in mind. But one of
these systems today, because it's very specialized, probably costs about
a million.
Speaker 1 (30:49):
Dollars to set up.
Speaker 2 (30:50):
Wow, that's another reason why you will probably won't carry
it around in your pocket any times soon. But it's
an important actually point to make, is that people will
build these systems and go to the larger scales. I
can and spend a lot of money to try to
do the first demonstrations, but we'll have to shrink them
and make the more cost effective all the components that
go in.
Speaker 1 (31:09):
Eventually, it's like we did for any algorithm.
Speaker 2 (31:11):
Yeah, exactly, and that part will happen. It just requires
you to start building these larger systems and for you know,
the companies that are making the individual components for them
to have larger volumes so they can drive down costs.
But where it's particularly challenging right now is actually in
the control electronics. Like the costs about maybe ten thousand
dollars a little more than ten thousand dollars just for
the control used for every single fubit. Wow, and we
(31:34):
need to go to maybe a million fubits or something.
So that's like ten billion dollars just in the control
hardware right if we were to scale out what we
have today, So it's very costly to imagine doing that,
so right now, yeah, but then we'll get better. We'll
do custom silicon chips, where the costs are in the
scale of the control electronics is much more efficient, so
we'll do what are called ASEX or custom circuits that'll
(31:54):
drive down costs tremendously. But yeah, that that has to happen,
but it just you know, it's not we're not quite
there yet.
Speaker 1 (32:01):
So there you have it. You can build a quantum
computer in your garage right now for about a million dollars,
although for that money right now you could only put
about twenty cubits on it, which is about as sophisticated
as an abacus, although this case would be a quantum ebicus. Right.
The last thing we'll talk about is why quantum computers
(32:21):
are so hard to make. If they can break any
encryption on the planet or potentially let us simulate new
chemicals and materials, why haven't we done it? What is
so hard about making a quantum computer? Here's Oscar explaining it.
Speaker 2 (32:35):
Probably the thing that makes it most difficult, and maybe
it's the most relevant to talk about, is that let's
say you want to do a computation with a quantum computer,
and you want to describe it by a certain number
of particles, and you want to use those particles to
do your quantum simulation. Then you need to be able
to control those particles, to manipulate them to do the
computation you want. But if those particles interact with the environment,
(32:58):
then part of the information that you wanted to control
or manipulate will actually evolve and become connected to these
other particles. And that's the really tricky problem is how
do I control tiny little quantum particles with my grubby
little hands, so to speak. So I have to be
able to send in these control signals to midp and
manipulate these quantum particles. But I can't let in any
(33:21):
other parts of the environment at the same time, and
so it becomes a really hard problem to sort of
shield the system you're trying to use to do this computation,
but then also allow yourself these control knobs.
Speaker 1 (33:34):
Is it like a question of purity to some.
Speaker 2 (33:37):
Degree, yes, Like the properties electron have to be just
that electron, and they interact with other things that you're
not able to control, you lose the information, all right.
Speaker 1 (33:47):
So the reason that quantum computers are so hard to
make and run basically goes back to Schrodinger's cat. Might've
heard of this analogy when people are talking about quantum things,
and the idea is that if I take a cat
and I put it inside a box, and I also
put in the box a quantum particle that might kill
the cat, then when I close the box, eventually the
(34:10):
cat becomes both alive and dead at the same time.
And that's because when I close the box, the quantumness
of that killer particle basically extends to the cat itself. Now,
a quantum computer is basically like taking a whole bunch
of those boxes with cats that are alive and dead
at the same time, and it tries to do math
(34:32):
with them. And because all those cats are in that
magical quantum state of being two different things at the
same time, alive and dead, then you can do some
really powerful computations with them, like multiply a whole bunch
of numbers all at the same time. But as soon
as anyone takes a peek inside one of those boxes,
then the whole thing collapses. As soon as you open
(34:52):
one box and you see whether the cat is alive
or dead, then that box loses its quantum magic, and
all the other boxes is that are talking to it
will also lose their quantum magic. So the reason you
need to build giant refrigerators and keep these computers in
an almost perfect vacuum with perfect coldness is to protect
them from any random bit of motion or energy from
(35:15):
essentially peeking inside your quantum boxes, because if that happens,
the whole thing collapses and stops working. And this problem
only gets worse as you make the computers bigger and
more complicated. But people like Oscar are getting better and
better at it. Well, that was great, That was awesome.
I guess just the last question, what is the current
(35:36):
state of the art in quantum computers?
Speaker 2 (35:38):
Yeah, so I think that if you can look at
this song, I would say three axis, So you can
ask how many physical cubits can I make in control
right now? Right?
Speaker 1 (35:46):
What's the highest number somebody has been.
Speaker 2 (35:48):
So if you just said I just want to be
able to control this many cubits, it's a few hundred.
And people have made systems of more than a few thousand,
but maybe not controlled all of them simultaneously. But people
have definitely made a few hundred and controlled them. So
we're getting to that level. And you might say, well, okay,
put that in context, and if we could control them
(36:09):
with high enough fidelity and not make errors, we would
be at the point where we could actually start to
access and solve problems of practical utility better than we
think other computers can. Like we could answer some of
these questions about how electrons interacted materials, like small toy problems,
but still useful.
Speaker 1 (36:25):
So like if I have a thousand cubits working, Yeah,
what kinds of passwords can I break? Right now? Yeah?
Speaker 2 (36:31):
So, like the number of bits and an rsa key
is like a few thousand, So if I had a
few thousand cubits, I could crack RSA.
Speaker 1 (36:39):
A few thousand, and we're at one thousand now. Yeah,
so right now we could maybe crack simple passwords.
Speaker 2 (36:45):
Like yeah, that's right, short short ones that we can
already do classically, so probably not useful, but we're within
striking distance. But the bigger problem is that we can't
do those calculations because our calculations are too air prone.
Then we need to add the air correction. Okay, that's
the other that's the and that's adding redundancy. And so
really think about this. I need to not have just
a few thousand physical cubits, but I might need a
(37:07):
few million, because the redundancy factor is pretty large right now.
Like if my hardware had no errors, I wouldn't need
to do any air correction, and there's redundancy factors one.
But I do have errors, and the errors we have
right now require about another factor of thousand overhead. A
thousand cubits multiply thousands of times, so it would be
a thousand tuns of thousand, which is a million. If
(37:28):
I need a thousand cubits do computations with I have
to multiply that by one thousand, and that gives me
how many physical cubits I need to represent? Oh wow,
So that's why I'm saying we probably needed like a
million physical cubits. So that's what people are doing right now.
The fact is that we can actually build and control
on order a few hundred one thousand cubits is amazing, right,
that's huge progress.
Speaker 1 (37:48):
Like ten years ago it was zero cubits.
Speaker 2 (37:50):
I would say we became masters of the individual cubit
so to speak. Maybe even in two thousand, we're really
really good at that. It was very hard to first
even figure like to control a single cubit. But since
then we've been already growing small cubit systems and improving
how the interacting in the gates that we can implement.
There is a recent result where scientists at Google showed
(38:11):
that their processor would require ten out twenty years for
a classic computer to simulate what they've done their processor.
You know, our own team Amazon, we focused on a
slightly different hardware implementation that potentially has an ability to
reduce the hardware overhead by factors on the order five
to ten, which could be very important. So even though
(38:33):
it doesn't have a practical application. Yet it's clear like
there's a big difference in the power of what these
things can do. There are a set of problems that
the classic computers are just not going to be good at,
and there's going to be a set of things that
quantic computers can do that classical ones cannot mimic. And
if you're watching this as a sort of an interested
techy observer and looking for a turning point or a
tipping point, I'd be watching for how these air rates
(38:54):
go down, how efficient air correction is in these sort
of one hundred two thousand cubit systems over the next
few years.
Speaker 1 (39:00):
Very cool, Well, thank you so much, Oscar. That was fantastic.
Speaker 2 (39:04):
Yeah, I hope we got into enough of the detail
where it's understandable enough. It is definitely a difficult subject,
and there's a lot of hype around it. Even for me,
it's very hard to read the news and to decipher
what is really in advance of what isn't. And I'm
deep in the field, so I can only imagine for
others that read about it.
Speaker 1 (39:21):
Very cool, right, all right, thanks a lot, yep, And
that is how a quantum computer works. Thanks for going
on this field trip. With me. I hope you enjoyed that.
See you next time. You've been listening to Science Stuff
production of iHeartRadio, written and produced by me or Hitchm
executive producer Jerry Rowland, an audio engineer and mixer Casey Pegram,
(39:45):
and you can follow me on social media. Just search
for PhD comics and the name of your favorite platform.
Be sure to subscribe to Sign Stuff on the iHeartRadio app,
Apple Podcasts, or wherever you get your podcasts, and please
tell your friends we'll be back next Wednesday with another episode.