I’m going way off what I would normally blog about for this one, but this has absolutely fascinated me. ‘Time Crystals’ were theorized back in 2012 and identified in 2016. They’ve really hit the news in the last few days as Scientists believe they have used Google’s Quantum computer to ‘create’ a Time Crystal.
But what is a Time Crystal and why would you care? well here’s my very basic overview, please bear in mind I am not a quantum physicist or scientist, just a very interested onlooker trying to make sense of it, so here goes…
A normal crystal is made up of a group of atoms which at normal temperatures are constantly spinning, they are unstable as you never know which direction they are pointing in. So this has limited value to us from a computing perspective.
Now if you cool these atoms down to a really really low temperature they stop spinning, they fix themselves to point in one direction only, they aren’t all necessarily pointing in the same direction though but they tend to line up meaning that they are pointing in one direction or 180 degrees in the other direction. This is effectively the second law of thermodynamics in action, at really low temperature there is very little energy available and therefore the atoms stop moving and line up as this is the lowest energy state. This is already interesting because if you know which direction they are pointing in you could theoretically assign that to a zero or a one which is how we store data today, as we know all data is just a collection of zero’s and ones.
But what makes a time crystal? Well if I hit the atoms with a laser pulse that has energy of a consistent frequency, the atoms start flipping and they do this in a consistent and predictable way so now we know what state they are in, we can see what direction they are pointing in and know when they’ll flip. But we already know this and that’s not what makes a time crystal, when I remove the laser they continue to move in exactly the same way, we’ve put them into a consistent and predictable state (Think of a perpetual motion machine), but again this is still not new.
Here’s the really cool part
Back to that second law of thermodynamics, once you remove energy from a system then over the time the system will slow down as it needs additional energy to keep moving, that’s why a perpetual motion machine is not really a perpetual motion machine, it’s always slowing down after that first injection of energy, or it needs to consume more energy in order to keep going.
Time Crystals don’t slow down!
Once you’ve hit them with the energy frequency they continue in this state of perpetual motion without needing any additional energy to sustain it. It also seems to be the case that if you hit them with a new laser with completely different energy frequencies they are not affected, they remember the original state that they were put in and keep consistently flipping in the same predictable way.
The thoery is that we’ve created a new type of matter, we’ve taken matter that was made up of atoms that are unpredictable as they are always spinning and have created matter made up of atoms that are moving in a consistent and predictable way and will keep moving this way forever without the need for any additional energy.
This goes against the second law of thermodynamics! now you’re starting to see why this is such a big deal.
How do Google believe they’ve proved this with their Quantum Computer?
(This has still yet to be peer reviewed), Google used 20 of the qubits on their Sycamore Quantum Computer (a qubit is a controllable quantum partical) it’s the building block for how Quantum computers work. Now think of these qubits in the same way as the atoms I’ve been describing above. What they’ve essentially proved is that they could mirror these qubits using a laser without actually consuming any net energy from the laser.
Okay that’s probably a bit heavy, but one of the biggest challenges in quantum computing is dealing with the state of qubits, they can be in two states simultaneously (think Schrodingers cat) so if this new discovery provides some way of manipulating the state, or even storing or managing the state then it could be a really really big deal.
It also could prove a real world use case for Quantum computers, I always think of computers as very very fast calculators, but does this mean that we can use qubits to actually simulate the creation of new materials at the atomic level? that takes our definition of what a computer is and utterly blows it apart.
Time crystals is a pretty misleading term for what this is, but there’s a lot of chatter now as to whether this is potentially a breakthrough in the world of Quantum computers, I’m sure that many of you probably think that the real world use of Quantum is also a long way into the future, so why would you care about these science projects but Gartner predict that up to 90% of organisations will be using Quantum by 2023, so watch this space.
It could also be possible that this has no value at all, nobody really knows at this stage.
Thanks for calling attention to this topic. I’ve been interested in entanglement for years, but have yet to see how any Qubit-powered device can create really meaningful insight. Yet.
The process you describe was also discussed in Quanta Magazine (https://www.quantamagazine.org/first-time-crystal-built-using-googles-quantum-computer-20210730) and one sentence in that article stands out: “Perhaps the main advantage of the [Google] machine over its competitors is its ability to tune the strengths of interactions between its qubits.” This means that interference due to a varying (random) logical set of interactions can achieve a locked-state synchronous behavior. But it doesn’t mean the calculations are determinative – it just means that the oscillations (flips) are present as an allowed state within the Qubit matrix. There’s a very strong correlation here between the way we think about the various synchronous Multi-Body state mechanisms: mechanical (pendulum, metronome, etc.), chemical (B-Z reaction) and others, and the way we think about quantum computers with thousands or millions of Qubits. Because of the limited way we tend to think about state in computing (binary and linear), small entangled quantum machines, even with adjustable interaction strengths (which make them non-determinative on the bit level) are kind of a toy. But it does seem that the types of problems these systems can solve are becoming clearer – and they are mostly in the fuzzy logic, AI space where systems cannot be broken down into smaller sub-problems.
Now if we could just find a way to program them… 🙂
Fantastic find! Thanks!
It’s a topic I’ve been trying to understand for a while, the example of time crystals was my aha moment. It suddenly made me think about how we can use qubits to simulate theories rather than thinking of it from the perspective of calculations.
Glad you enjoyed the post!
Pretty wicked to imagine we may be reading about the next wave of something potentially important. Fun read.
Thanks Luis, the whole topic of quantum is vast right now and it’s amazing that we’re already finding use cases. We may be a long way off creating a true quantum computer with enough qubits and error correction to solve problems of any meaningful size but, companies are already creating quantum annealers and are creating new quantum algorithms, or even using ones that have already been created, to solve real world problems right now.