00:01What do we want? Better memory.
00:03When do we want it? Want what?
Joking aside, I have a two Terabyte
00:08hard disk and used to think that’s a lot. Then
I bought a new video camera and it’s clogged my
00:14disk in no time. And yes there’s cloud storage
but you know, it’s not actually in the clouds.
00:20They still store it on some thing down on earth.
And besides that, it’d take me forever to upload.
00:27So I was thrilled to see this recent
new paper in Nature about a new optical
00:32storage method that could bring disk memory
into the Petabyte range. Petabytes. That’s
00:39a thousand Terabytes. Let’s have a look.
The new technology is an old technology,
00:45it’s optical storage. Some of you may even
be old enough to remember optical storage.
00:51We used to carry music around on those things
called compact disks and let me just say you
00:57had to walk very carefully if you didn’t want
to listen to the same sentence a thousand times.
01:03Later we put movies on these things called DVDs.
01:08We have a neighbour who’s hung her old DVDs up on
the balcony to scare off the pigeons. True story.
01:15These compact disks were usually made of some
kind of plastic with a coating, which gave them
01:21their shiny appearance. They worked by using a
laser to burn the information into the coating,
01:28and then read it out again. Actually, a
pretty nice idea, quite straight forward.
01:34Why did compact disks die? They
were outcompeted by hard drives
01:38that can store information packed much more
densely. The information density on compact
01:44disks or any optical storage is ultimately
limited by the frequency of the laser light,
01:50for visible light that’s a few hundred nanometres.
In modern flash drives the information is stored
01:56in little magnetizable cells that are just a
few atoms in size and can be as small as 10
02:02nanometers or so. They can store much more data in
the same space, hence the death of compact disks.
02:09But maybe they’ll soon make a comeback. Because
in this new work a group of researchers from China
02:15figured out how to write data in multiple layers
on a disk with lasers and read them out again.
02:22They say this could work up to hundreds of
layers, which would be a huge increase in
02:27memory capacity. They say that a single disk of
this type could store as much as a Petabit. Yes,
02:34a bit isn’t the same as a byte, but
then this was only a prototype. They
02:38also envision that in data centres the disks
could further be stacked into arrays that
02:44could each hold exabits. Sounds good
but just exactly what did they do?
02:50This new work is really a combination
of several advances. Partly they made
02:55it work because they have a synthetic
material that can be modified in a very
03:00targeted way by changing how its polymers
are linked. They can write at a certain
03:05depth in the material by focusing
the laser on one particular spot.
03:10They actually use two lasers, one that initiates
a change of the molecular property of the disk,
03:17the second one that terminates it, leaving behind
an altered region. This region can later be read
03:23out by causing it to emit light with yet another
laser and that can then be detected. So, to write,
03:29you locally change the property of the material
with lasers. To read, you cause it to emit light.
03:36The other advancement that they make
use of is that this two laser method
03:40also allows them to encode structures below the
03:43wave-length of the lasers. This also makes
it possible to pack the data more densely.
03:49I guess it also helps that they were
working with femtosecond that prevents
03:53them from melting the entire disk to a pool
of goo before they’re done working with it.
03:58Writing to and reading data from
the disk actually works very nicely,
04:03as you can see in this video. What you
see here is a scan through the layers.
04:08They have alternatingly encoded the acronyms
of their institute and the university and you
04:13can see that they are pretty clearly
readable. The distance between two
04:18layers is just about 1 micrometre. So you
could pack 1000 of them into a millimetre.
04:25That said, there is some way to go from there
to commercialization. First and most importantly
04:30when it comes to storing data it’s not just
the density of the storage that matters,
04:36it also matters how quickly you can write
the data and and read it out. They didn’t
04:41say anything about that, but I assume since this
is a prototype it’s fairly slow. Then there’s the
04:47question how much energy that takes, and also,
where do you get a femtosecond laser from.
04:53Then again, you know, this entire area of
technology is currently evolving so quickly,
04:58these challenges might be possible to
overcome given a few years’ time. So who
05:03knows maybe compact disks will make a comeback
in the not too far future. And inevitably the
05:09day will come when we’ll all record our
entire life in 8K, I can’t wait for it.
05:15Watching science videos is all well and fine and
I don't need to complain you're doing it. But it's
05:21not a good way to actually learn something
new. If you want to learn something new,
05:26you need to actively engage with the topic. A free
and easy way to do this is on Brilliant.org who've
05:33been sponsoring this video. On Brilliant, you find
courses on a large variety of topics in science,
05:39computer science, and mathematics. It's a fresh
and new approach to learning that makes growing
05:45your knowledge easy and fun. I've learned so much
there. All their courses come with interactive
05:51visualizations and follow up questions. Some
also have videos for demonstration experiments or
05:57executable Python scripts. This really gives you
a feeling for what's going on. Whether you want
06:03to know more about solar panels, neural networks,
astrophysics, special relativity, or computational
06:10biology - Brilliant has you covered. I even have
my own course there that's an introduction to
06:16Quantum Mechanics. It's a beginners course
and covers topics such an interference,
06:21super positions and entanglement, the uncertainty
principle, and Bell's theorem. And after that, you
06:27can continue learning more about quantum objects
or maybe quantum computing. To give it a try
06:33yourself, use our link brilliant.org/sabine and
sign up for free. You'll get access to everything
06:39Brilliant has to offer for 30 days, and the first
200 subscribers using this link will also get 20%
06:46off the annual premium subscription.
Thanks for watching, see you tomorrow.
