Previous: Is Running Really Bad for Your Knees?
Next: Why Is Fake Blood so Hard to Make?



View count:446,164
Last sync:2022-12-27 04:45
Today's data storage solutions have an expiration date. What's on the horizon to replace them?

Hosted by: Michael Aranda

SciShow has a spinoff podcast! It's called SciShow Tangents. Check it out at
Support SciShow by becoming a patron on Patreon:
Dooblydoo thanks go to the following Patreon supporters—we couldn't make SciShow without them! Shout out to Kevin, Bealer, Mark Terrio-Cameron, KatieMarie Magnone, Patrick Merrithew, Charles Southerland, Fatima Iqbal, Sultan Alkhulaifi, Tim Curwick, Scott Satovsky Jr, Philippe von Bergen, Bella Nash, Bryce Daifuku, Chris Peters, Patrick D. Ashmore, Piya Shedden, Charles George
Looking for SciShow elsewhere on the internet?
Computers are getting smaller, faster, and more powerful all the time.

That’s awesome for a lot of reasons, but it’s creating a problem: most of our computers store data pretty much the same way they have for decades, and that technology is starting to run into fundamental physical limits on just how small and fast it can be. Which is why a lot of researchers are working on newer tech to take its place.

And someday, you might be saving that fanfiction you’re writing about Sherlock and Watson to a tiny glass disk that could last billions of years, or storing it as a hologram. You might even be able to hard-code it into DNA. You probably know that files on your computer are stored as lists of ones and zeros.

But we haven’t really talked about how those ones and zeros are actually stored in your computer. The usual explanation is that ones are ‘on’ and zeros are ‘off’, but that doesn’t explain how computers keep track of them without power. Nothing’s really on when you shut down your computer, but all your songs and documents stay saved anyway.

That’s because computers don’t permanently store data in patterns of ‘on’ and ‘off’. Instead, ones and zeros, called bits, are stored in different ways, like in magnetic patterns on a hard disk drive. Computers read a hard drive’s data by spinning the disks to the right place and reading the pattern with a tiny magnetic head.

It’s kind of like how old record players worked, but the head is way smaller and hovers above the disk instead of rubbing against it. And you might remember floppy disks, which were a portable version of this technology. We’ve been permanently storing data on hard drives for decades, since the days when a file the size of a song would fill most of a computer the size of a fridge.

Today’s hard drives only need a few hundredths of a millionth of a square meter to hold thirty million or so bits in an average song, but the technology is still pretty similar, just smaller. We keep using these hard drives because they’re pretty stable. Data stored as magnetic patterns generally stays that way for years, if not decades.

But those stored bits keep getting smaller and smaller, and there are limits to how small bits can get before they start changing the bits around them. Spinning hard disks at thousands of revolutions per minute also takes a lot of energy, and all those rapidly moving parts can break or wear out. Optical disks like CDs and DVDs also store data in patterns, although they use physical patterns of bumps on their surfaces, which computers read by bouncing a laser off the bumps.

But there’s still a limit to how many bumps you can cram together, since each kind of laser only reads bumps above a certain size, and smaller bumps generally need more expensive lasers. Computers can also store data with transistors, which are basically little switches where the ones and zeros really are ‘ons’ and ‘offs’. When you open a file, for instance, it gets copied over to random access memory, or RAM.

RAM is made of transistors that either block electric current for a ‘zero’ or let it through for a ‘one’. Transistors don’t have any moving parts, so they can quickly change between different states. It’s also a lot faster to read and write data to transistors, since there’s no spinning disk involved.

But you can’t use RAM for permanent storage because without power, the transistors are reset to the ‘off’ position. Instead, you can permanently store data on solid state drives, or SSDs, which use different kinds of transistors that don’t need constant power to store data. That’s because they can let a charge build up and get stuck in parts of the transistors.

A transistor with a charge represents a “zero”, and a transistor with no charge is a “one”. And the charges stay stuck even when there’s no power. Another advantage of SSDs is they don’t have any spinning disks or other moving parts that will break.

SSDs are still much less common than traditional hard drives, but they’ve become more popular over the last few years as they’ve gotten bigger and cheaper. They’re faster, and even though the transistors in SSDs can wear out if they’ve been used enough times, they’ll usually take longer to wear down than it’ll take you to replace your computer. SSDs have seemed like the wave of the future for the past few years, but someday they might be as obsolete as the floppy disk.

Because scientists are working on completely different ways of storing data. In 2013, a team at the University of Southampton in the UK came up with what they’re calling five-dimensional data storage. They’re thumb-sized disks with patterns etched into them, a lot like how CDs and DVDs have data imprinted on their surfaces.

But CDs, like most modern data storage technologies, mostly just store information in two dimensions. DVDs can do a little better, since they can actually have two different patterns, one on the surface, and one slightly underneath it. But these disks take that to another level.

They have patterns cut into them with ultra-fast lasers in three different layers, and each layer has two different patterns in it. So lasers reading these disks can focus in one of five different ways, and each way they’ll read completely different information. This is where the “five-dimensional” part comes from.

With all of these dimensions available, the researchers estimate that each disk can hold 360 terabytes of data. That’s about three quadrillion bits, or enough to store the entire Library of Congress on fourteen little disks. Plus, these disks are made of glass, which is one of the most stable materials we have.

If we’re lucky, data on some of today’s permanent storage devices might last between a few years and a few decades. But glass can withstand really high temperatures and pressures, and it’s stable around lots of different chemicals. Thanks to that glass, data on these disks could stay intact for billions of years!

So computers of the future might come with these tiny glass disks inside them, and tons of storage space. Then there’s stuff that just sounds like science fiction. Take holographic storage.

It's called "holographic" because it uses holography, where the interference of light encodes data, and it would work a little bit like a CD or one of those 5D glass disks: To read something in holographic storage, you’d shine a laser on something with a pattern in it. But there are a couple big differences. For one thing, there wouldn’t just be one or three patterned layers; there might be thousands.

The laser would go through whatever crystal or other material had the pattern, instead of bouncing off of it, so it could be focused on one layer after another throughout the entire thing. It also wouldn’t have to read one bump or scratch at a time, like you have to with CDs or glass disks, or even hard drives and SSDs. Instead, the laser would shine through the crystal onto something like a camera, which would capture the pattern of the entire layer at once.

So instead of reading one bit after another like our computers do today, computers with holographic storage might be able to read sixty thousand bits at a time. But our computers mainly work by analyzing one bit at a time, which means we’d have to rework the way that our computers themselves approach information. So it’s still far off in the future, but holographic storage is in the works.

But maybe you want permanent storage that feels a little more personal than eternal glass disks or patterns in crystals. Well, you’re in luck. Your DNA is made of chemical compounds called nucleotides that tell your body what kinds of molecules to make.

And scientists have been working on ways of arranging those nucleotides to encode data that computers can work with. Nucleotides are smaller than the smallest magnetic bits or transistors that computers use to store data today. So if each nucleotide in a strand of DNA represented one bit of information, DNA could be way more efficient than anything else in the world right now.

We’re talking storing the entire world’s data in just a teaspoon of DNA. And DNA could also be more stable than a lot of other current methods we use. It’s not age-of-the-universe stable, but it might be able to last for hundreds of years longer than hard drives or SSDs.

DNA storage has only been around for about twenty years, so it hasn’t quite reached its potential yet. Scientists are still figuring out how to get these incredibly tiny nucleotides in exactly the right order along an entire strand of DNA so that each one can represent a bit of information. The current record is from a team that stored about two hundred megabytes (about 1.6 billion bits, or about sixty songs’ worth of data) on short DNA strands.

It’s also hard to find ways of reading DNA at a particular random spot, which is what computers have to do whenever they open a file. If researchers work out these problems, though, there could be a day when you really do have music in your DNA. Or, at least, in your computer’s DNA.

For more in-depth science behind the technology that runs our world, check out our recent mini-series on the history of the internet. And if you’re new to SciShow, don’t forget to subscribe!